evaluation of delegated cooperation desk report · evaluation of the eu aid delivery mechanism of...

1

Evaluation of the EU aid delivery mechanism of delegated cooperation 2007-2014

International Cooperation

and Development

EuropeAid

Evaluation of the EU aid delivery mechanism of delegated cooperation

(2007-2014)

Final report Volume 3 – Evaluation matrices

November 2016

______________ Evaluation carried out on behalf of

the European Commission

2


Consortium composed by

ECDPM, Ecorys, Lattanzio, Mokoro and Particip

Leader of the Consortium: Ecorys

Contact Person: Albert de Groot

[email protected]

Contract No EVA 2011

Specific contract N° 2014/343011

This evaluation was commissioned by

the Evaluation Unit of the

Directorate General for International Cooperation and Development –

(European Commission)

The opinions expressed in this document represent the authors’ points of view

which are not necessarily shared by the European Commission

or by the authorities of the concerned countries.

The evaluation is being managed by the DG DEVCO Evaluation Unit

3


This report has been prepared by:

ECORYS Nederland B.V. Watermanweg 44

3067 GG Rotterdam

P.O. Box 4175 3006 AD Rotterdam

The Netherlands

T +31 (0)10 453 88 00 F +31 (0)10 453 07 38

E [email protected] Registration no. 24316726

W: www.ecorys.nl

Evaluation team:

Albert de Groot * Martin van der Linde *

Ivo Gijsberts * Karen McHugh Anneke Slob

Andrea Dijkstra * Dafina Dimitrova * Jonathan Wolsey *

* Staff of ECORYS

Rotterdam, November 2016

4


5


Table of contents

List of abbreviations 7

EQ 1: Transaction costs 9 3.13.1.1 Definition and scope of the EQ 9 3.1.2 Rationale of the EQ 10 3.1.3 Methodological observations and challenges 11 3.1.4 Improved division of labour (JC 1.1) 12 3.1.5 More co-financing (JC 1.2) 17 3.1.6 Larger programmes (JC 1.3) 21 3.1.7 Use of single management systems (JC 1.4) 26 3.1.8 Reduction of number of active donors (JC 1.5) 29 3.1.9 Limited additional transaction costs (JC 1.6) 31

3.2 EQ 2: Ownership and leadership 35 3.2.1 Definition and scope 35 3.2.2 Rationale of the EQ 36 3.2.3 Methodological observations 36 3.2.4 Formulation of policies and strategies (JC 2.1) 37 3.2.5 Implementation & monitoring of policies (JC 2.2 & 2.3) 39 3.2.6 Reduction of number of active donors (JC 2.4) 43

3.3 EQ 3: Complementarity and Added Value 46 3.3.1 Definition and scope of the EQ 46 3.3.2 Rationale of the EQ 46 3.3.3 Methodological observations 47 3.3.4 Improved division of labour (JC 3.1) 48 3.3.5 Increased use of comparative advantages (JC 3.2) 50 3.3.6 Improved donor coordination and harmonisation (JC 3.3) 54

3.4 EQ 4: Aid Fragmentation 58 3.4.1 Definition and scope of the EQ 58 3.4.2 Rationale of the EQ 59 3.4.3 Methodological observations and challenges 59 3.4.4 Improved division of labour (JC 4.1) 60 3.4.5 More co-financing (JC 4.2) 61 3.4.6 Larger programmes (JC 4.3) 62 3.4.7 Improved donor coordination and harmonisation (JC 4.4) 63

3.5 EQ 5: Alignment 66 3.5.1 Definition and scope of the EQ 66 3.5.2 Rationale of the EQ 67 3.5.3 Methodological observations 67 3.5.4 Embedded in policies and strategies of partner country (JC 5.1) 68 3.5.5 Based on national procedures and systems (JC 5.2) 69

3.6 EQ 6: Visibility 73 3.6.1 Definition and scope of the EQ 73 3.6.2 Rationale of the EQ 73 3.6.3 Methodological observations 74 3.6.4 Visibility of the EU in Delegation Agreements (JC 6.1) 75 3.6.5 Visibility of the delegating donors in Transfer Agreements (JC 6.2) 78

3.7 EQ 7: Balance of DAs and TAs 81

6


3.7.1 Definition and scope of the EQ 81 3.7.2 Rationale of the EQ 81 3.7.3 Methodological observations 82 3.7.4 Inventory of TA/DA ratios 82 3.7.5 Constraints at the level of the EU (JC 7.1) 84 3.7.6 Constraints at the level of the (potential) TA partners (JC 7.2) 85 3.7.7 DAs are much more attractive than TAs (JC 7.3) 86

3.8 EQ 8: Assessment of DC proposals 88 3.8.1 Definition and scope of the EQ 88 3.8.2 Rationale of the EQ 88 3.8.3 Methodological observations 89 3.8.4 Motivation of EUD to enter into DC agreements (JC 8.1) 89 3.8.5 Appropriate format of Assessment Fiches (JC 8.2) 93 3.8.6 Quality of DA Assessment and Action Fiches (JC 8.3) 95 3.8.7 Quality of TA Assessment Fiches (JC 8.4) 97

3.9 EQ 9: Quality of cooperation 99 3.9.1 Definition and scope of the EQ 99 3.9.2 Rationale of the EQ 99 3.9.3 Methodological observations 99 3.9.4 Cooperation on the basis of clear rules and procedures (JC 9.1) 101 3.9.5 Provision of timely and adequate information (JC 9.2) 104 3.9.6 Quality of the coordination between the partners (JC 9.3, 9.4 & 9.5) 107

7


List of abbreviations

AAP Annual Action Programme AECID Agencia Espanola de Cooperacion Internacional al Desarrollo AIDCO Directorate general of the EC charged with development cooperation (until

2010) AFD Agence Française de Développement APONTRA Programme d’Appui à la Politique Nationale des Transports BMZ Bundesministerium für wirtschaftliche Zusammenarbeit und Entwicklung BTC Belgian Technical Cooperation CICL Camoes Instituto da Cooperacao e da Lingua CNCS National AIDS Council (Mozambique) CRIS Common Relex Information System CSP Country strategy paper (EU) DA Delegation Agreement DAC Development Assistance Committee DANIDA Danish agency for international development DC Delegated cooperation DEVCO Directorate general of the EC charged with development cooperation DFID UK Department for International Development DG Directorate General EDF European Development Fund ESPS Environmental Sector Programme Support (Mozambique) EU European Union EUD EU Delegation EQ Evaluation question FR Financial Regulations GCCA Global Climate Change Alliance GBS General budget support GIZ Deutsche Gesellschaft für Internationale Zusammenarbeit HQ Headquarter IED Intended Effects Diagram IMDA Indirect Management Delegation Agreement JAS Joint Assistance Strategy JC Judgement criterion KfW Kreditanstalt fur Wiederaufbau MoU Memorandum of Understanding MS EU Member State M&E Monitoring and evaluation NGO Non-governmental organisation NIP National indicative programme OCT Overseas countries and territories OECD Organisation of Economic Cooperation and Development PAFIRIZ Projet d'appui à la Filière Riz au Bénin PAOSC Programme d'appui aux organisations de la société civile (Mali) PARAQ Programme d’appui à la reconstruction et à l’aménagement de quartiers

pour aciliter le retour des populations sinistrées (Haiti) PARADDER Programme d’appui à la réforme administrative, à la décentralisation et au

développement économique regional (Mali) PEGASE Mécanisme Palestino-Européen de Gestion de l'Aide Socio-Economique

(Palestine) PRODEL Program for Local Economic Development RCA Revealed Comparative Advantage SC Steering Committee

8


STAR Strengthening Transparency, Accountability and Responsiveness (Ghana) TA Transfer Agreement TECNICA Programa de Apoyo a la Educación Técnica y Formación Profesional en

Nicaragua TVET Technical and Vocational Education and Training UN United Nations

9


EQ 1: Transaction costs 3.1

EQ1: To what extent has Delegated Cooperation reduced transaction costs?

3.1.1 Definition and scope of the EQ

This Evaluation Question (EQ) assesses to what extent Delegated Cooperation (DC) has reduced transaction costs during the formulation, implementation and monitoring of the DC project or programme. A generally used definition of transaction costs is presented on the aid effectiveness website and reads as follows: “All costs involved in the negotiation, management and administration of development aid, on both donor and recipient sides”1. For the purpose of this evaluation, this overall definition will be interpreted in the following way: transaction costs include all costs at the level of the donor and the partner country related to the preparation, identification, formulation, management, monitoring and evaluation of the aid-funded project or programme, excluding the costs actually funded by the project or programme itself. Several studies and papers indicate that it is very complicated to assess transaction costs2. Lawson mentions that so far (up to 2009) only two studies had made a serious attempt to measure aid transaction costs, but failed to achieve that objective. That is why he proposed that the impact of the Paris Declaration on transaction costs should be evaluated in relative and qualitative terms (perceptions of stakeholders about increases or decreases of transaction costs) without attempting to quantify transaction costs3. In the final report of that evaluation (annex 5, p. 209), it was mentioned that “no internationally agreed definition exists”. The term is “unfamiliar to many, misapplied in the view of economists and generally an unhelpful concept for further analysis. Instead … the concept of the respective burdens of aid management was integrated across the evaluation matrix”. Despite these methodological challenges, the final report (2011) of the Evaluation of the Paris Declaration concludes that the aggregate transaction costs of aid are not falling4. A second message delivered by that study is that the implementation of the Paris Declaration itself has generated additional transaction costs both for partner country governments and donor agencies in the form of working on common approaches, establishing multi-donor structures and setting up new coordination mechanisms. Meanwhile, it was not clear whether those additional transaction costs were only transitional and whether those costs could be reduced in the medium to long term. Also the three monitoring reports of the EU Fast-Track Initiative on Division of Labour concluded that transaction costs have increased for both partner countries and donors as a result of an increased number of donor meetings linked to Division of Labour5. On the other hand, some authors have observed that transaction costs related to the division of labour have tended to rise during an initial phase as the framework is set up, but decreased thereafter because each donor was active in a lower number of sectors than before6. Other studies that analysed the operational effectiveness of basket funding arrangements (a co-financing arrangement) are not overly positive in terms of the

1 www.aideffectiveness.org, See under Tools and Definitions.

2 Andrew Lawson, November 2009, Evaluating the transaction costs of implementing the Paris Declaration, p.10.

3 Ibid, pp. 12-14 and 22-23.

4 The Evaluation of the Paris Declaration, Final Report, Copenhagen, May 2011.

5 EU (2011): EU accountability report 2011 on financing for development : review of progress of the EU and its member

states, Brussels. 6 Weingärtner, L. (2008): International workshop on rationalizing aid delivery : partner country experience and

perspectives, Pretoria: BMZ.

http://www.aideffectiveness.org/

10


transaction costs effects7. These studies revealed that the initial transaction costs of establishing a genuine multi-donor delivery channel (initial coordination & consensus building on management procedures) are particularly high. The link between DC and transaction costs has not been much looked into in the aid literature. A Norad paper published in 2006, surveying DC agreements among NORDIC+ countries, observed that DC might generate additional transaction costs for both donors and partner countries because DC arrangements have to be prepared and negotiated between donors and with the partner government8. However, once arrangements have been negotiated, they are expected to reduce transaction costs for the partner country, as the delegating donor is supposed to “remain silent” during project implementation. Given the conceptual and methodological problems to assess transaction costs in this evaluation, a simple straightforward approach has been adopted based on the reconstructed intended effects diagram in order to assess the contribution of DC to the transaction costs. The evidence is primarily based on the nine country case study supported by findings of the Global Climate Change Alliance (GCCA) case study, responses to the questionnaire and interviews with the Commission and EU Member States’ (MS) ministries and implementing agencies. Because of the difficulty to quantify transaction costs, emphasis has been put on collecting views and perceptions as expressed in documents and reports and by interviewing stakeholders.

3.1.2 Rationale of the EQ

Reducing transaction costs of both the partner countries and the donors, including the workload at the level of EU Delegations (EUDs), is an important envisaged outcome of DC. Reducing the workload of EUDs was identified by the 2007-2012 review of Delegated Cooperation as a major reason in practice for starting DC: “Delegated cooperation has been a means for division of labour within sectors, but decisions to use it have been to date largely guided by workload considerations9”. The same review underscored that DC should result in quantifiable efficiency gains and reduced transaction costs for partner countries. The Guidance paper on DC with Member States, also issued in 2012, further emphasized that DC should result in efficiency gains and reduced transaction costs for both the EU and partner countries. This link would be verified for each Delegation Agreement (DA) proposal, through an Assessment Fiche to be attached to the proposal for taking a Financing Decision as regards that DA10. Both documents, as well as other EU policy and strategy documents on DC, do however not elaborate extensively on how DC could or should contribute to reduction of transaction costs, neither at the level of the EUDs nor at the level of the partner country governments. Regarding the workload of EUDs, the Report on Delegated Cooperation 2007-2012 noted that “from the operational point of view, [DC] has allowed operational staff in Delegations to have more flexibility and make use of the expertise and experience of Member States' agencies, which is largely perceived less time consuming and demanding on Delegation resources than using other implementation modes, in particular when the latter would have involved procurement of contracts or managing work programmes/cost estimates”11.

7 See for example Common funds for sector support, building blocks or stumbling blocks, ODI Briefing Paper, February

2008. 8 NORAD (Norwegian Agency for Development Cooperation) (2006): Nordic plus practical guide to delegated

cooperation, Oslo. 9 Report on Delegated Cooperation 2007-2012, pp.16, ARES(2012) 388917, dd. 30.03.12.

10 However, DA Assessment Fiches have not been systematically elaborated: see EQ8.

11 Report on Delegated Cooperation 2007-2012, pp.16, ARES(2012) 388917, dd. 30.03.12.

11


These perceived advantages of DC may be particularly relevant in “cases where the Commission provides assistance to partner countries where it has no permanent office, such as in the smaller island states and OCTs in the Pacific and Caribbean, and where one or more Member States (agencies) may have a permanent representation”12. Regarding reduction of transaction costs of partner countries, the report argues that DC should result in larger programmes, which are supposed to bring along lower transaction costs (per euro of aid provided) than various smaller projects, thanks to fewer meetings, fewer reports and lower management cost per euro of aid provided and having to deal with less donors. While underscoring the benefits of these potential transaction cost decreases for both EUDs and partner countries, the same document notes that it is not possible to conclude whether the balance of all effects on workload and transaction costs of all players involved in DC had been positive or negative. On the one hand, a better division of labour, larger programmes and less aid fragmentation should have reduced the transaction costs, but on the other hand the six-pillar assessments and negotiations with the DA partner (the fund manager), additional monitoring and handling derogations cause additional work.

3.1.3 Methodological observations and challenges

Some main methodological challenges regarding the assessment of transaction costs have already been mentioned in section 3.1.1 and are not be repeated here. The methodological approach is outlined below. When analysing transaction costs, it is important to make a clear distinction between transaction costs at the level of the EU, the DC partners and the partner countries. Most DC policy and strategy documents refer to the overall potential of DC to reduce transaction costs of ‘all players involved in delegated cooperation’, without further specification of possible advantages for the EU, the DC partners and the partner countries. However, these advantages (or their absence) could differ a lot for each of these three categories, while the impact (either negatively or positively) could also differ. There could be differences in terms of workload reductions (or increases) but also in terms of the costs/benefits of the workload changes. Although the perceptions of the changes in transaction costs of the partner countries were taken into account during the field missions (for every country, between at least two up to ten government officials were interviewed), it was difficult to find substantial information pointing at an increase or decrease in transaction costs. Often it was found that DC had limited to no impact on the level of transaction costs incurred by the government. This was mainly because of the relatively low level of involvement of the government in DC-funded programme/projects. The only clear benefit appeared when the DC partner took a truly silent role, limiting the amount of stakeholders for the government to deal with. For the remainder of DC projects, it was found that most of the transaction costs were incurred or saved by the EU and the DC partner. Therefore, the analysis is primarily focused on the transaction costs of the EU and the DC partners. The analysis of the contribution of DC on transaction costs has been based on the contribution of each of the five related DC outputs to the observed change of transaction costs. According to the reconstructed Intended Effects Diagram (IED, Volume 1, figure 1.2) there are five (potential) DC outputs that could possibly contribute to a reduction of transaction costs, namely: improved division of labour, more co-financing, larger programmes, using single management systems and a reduced number of donors in the

12

Report on Delegated Cooperation 2007-2012, pp.15, ARES(2012) 388917, dd. 30.03.12.

12


sector. These outputs have been considered as Judgement Criteria (JCs). The links between these five outputs and the transaction costs at outcome level have not been substantiated further in the DC policy and strategy documents, but these links have been explored during this evaluation. In addition to the five outputs, two other related aspects have been taken into account for the assessment of transaction costs, namely the workload of both the EU and the DC partner. A separate JC has been devoted to explore possible additional transaction costs of DC.

3.1.4 Improved division of labour (JC 1.1)

JC 1.1. Delegated Cooperation has led to improved division of labour between the donors which has contributed to a reduction of transaction costs

This JC assesses whether DC has facilitated and promoted the implementation of division of labour principles, which was supposed to have contributed to a reduction of transaction costs. Division of labour is an important topic in the aid effectiveness and efficiency agenda and, according to the review of DC 2007-2012, it has been the main driver of Delegated Cooperation13. The OECD states that “development results can be improved when donors individually and collectively rationalise their activities at the country level”14. The second part of that quotation could be taken as a kind of definition of division of labour: ‘individual and collective rationalisation of donor activities at the level of the partner countries’. A better division of labour leads to a more rational allocation of aid and creates economies of scale, which results in higher complementarity, better alignment and a significant reduction of transaction costs15. As an envisaged DC output, division of labour is thus expected to contribute not only to reduction of transaction costs, but also to strengthening complementarity, increasing the added value and reducing aid fragmentation (all outcome indicators as shown in the IED, see figure 1.2 in Volume 1). Findings from the literature review Measuring division of labour is not easy, as there are many dimensions (division of labour internationally or within a country, or across sectors, or within large programmes or projects) and clear and measurable indicators have not been developed. The Compendium on Good Practices on Division of Labour concluded that it is not possible to capture progress made towards a better division of labour in one single indicator16. To do so, most studies look at trends in aid fragmentation (dispersion of aid between numerous donors and/or projects within a partner country) and donor proliferation (dispersion of aid between numerous sectors/areas from the perspective of a donor)17. Such studies faced several challenges such as time lags, intervening variables, incomplete data sets and varying sector definitions. Dreher and Michaelowa (2010) discussed and tested various indicators to measure division of labour, but concluded that there is not one indicator exhaustive enough to measure aid fragmentation or donor proliferation and that it is questionable whether quantitative indicators are the way

13

Guidance paper on Delegated Cooperation, 2012. 14

Evidence on Trends in Fragmentation and Proliferation and Implementation and Results of Division of Labour

Processes, Key Messages for HLF 4 and Beyond, TT DoL (WP-EFF, Cluster C) – May 4, 2011. 15

Bürcky, Trends in In-country Aid Fragmentation and Donor Proliferation, 2011. 16

Compendium on Good Practices on Division of Labour, Report prepared for the European Commission & OEDC DAC

(Input to the HLF-3 Meeting in Accra, September 2008), June 2008 (draft). 17

Axel Dreher and Katharina Michaelowa. Methodology to measure progress towards in-country division of labor, GTZ

study, November 14, 2010.

13


forward18. They stated that division of labour is very country-specific, which is among others caused by different sector definitions across countries. This evaluation therefore abstained from measuring division of labour on the basis of statistical formulas, but opted for a qualitative approach to assess how DC has contributed to applying the principles of division of labour. The review of DC 2007-2012 stated that DC was initially seen as an innovative tool for donors to streamline their participation in a specific country by focusing on a limited number of sectors19. In fact, DC was supposed to make it possible for donors to become a passive donor in their non-focal sectors by delegating fund-management to a donor, which had identified the relevant sector as a focal sector. However, the review concluded that DC has been used for that purpose (inter-sectoral division of labour) only in a limited way. DC was mainly used for supporting and strengthening lead donor arrangements within a sector and avoiding overlaps between donors within a sector by re-arranging the work and the projects between donors supporting a certain sector (intra-sector division of labour)20. Findings from the portfolio analysis Delegated Cooperation has had limited effect on improving inter-sectoral division of labour. It appeared that it was not a prerequisite to use DC only in an EU non-focal sector, to exit the sector or become a silent donor. This led to the result that in the entire portfolio, DAs and TAs have been signed in focal and non-focal sectors, as well as for global or regional programmes, where focal sectors are not applicable. For the entire portfolio of 164 DAs, the findings are as follows:

70 DAs are supporting projects in focal sectors of the EU;

64 DAs are supporting projects not being part of EU’s focal sectors;

30 DAs are part of global or regional programmes, or consist of ad-hoc support where the focal sector issue is not relevant.

These findings show that 39% of the DAs could have contributed to inter-sectoral division of labour: here the EU could have used DC to be a passive donor in a sector that does not belong to its focal areas. In the majority of the cases, the DA had little effect on adherence to focal sectors, as the EU was signing DAs in its own focal sector or for a regional or global programme. These cases could not have affected inter-sectoral division of labour. For the entire portfolio of 59 Transfer Agreements (TAs), the findings are as follows:

21 TAs are supporting projects in focal sectors of the EU;

18 TAs are supporting projects not being part of EU’s focal sectors;

20 DAs are part of global or regional programmes, or consist of ad-hoc support or General Budget Support where the focal sector issue is not relevant.

From this, it can be concluded that 36% of the TAs could have contributed to inter-sectoral division of labour, as for these TAs funds were delegated to the EU for an intervention in its focal sector. 18 TAs had no or possibly even a negative effect on inter-sectoral division of labour, as the EU received transfers for programmes in a non-focal sector. About one-third of the TAs were signed for global or regional programmes, or cross-sectoral support, and had no relation to a focal sector strategy. Findings from the case studies In the following table the scores regarding 46 DC projects and programmes are summarized, which shows DC did not change the inter-sectoral division of labour in 31 out of 46 projects and programmes (67%). In addition, seven projects showed a negative

18

Dreher and Michaelowa. Methodology to measure progress towards in-country division of labour, 2010. 19

Report on Delegated Cooperation 2007-2012, March 30, 2012. 20

Ibidem.

14


effect (15%) and eight showed a strong or modest positive effect (17%) of DC on the inter-sectoral division of labour. A main finding is therefore that DC did not have an important effect on improving inter-sectoral division of labour. Table 3.1.1 Case study scores: Effect of DC agreements on improving inter-sectoral division of labour

Country Strong effect

Modest effect

No change Negative

effect Total

Benin 0 1 6 0 7

Ghana 1 0 2 0 3

Mali 0 0 2 2 4

Mozambique 0 2 4 0 6

Nicaragua 0 0 2 2 4

Palestine 0 1 8 3 12

Tanzania 2 0 1 0 3

Timor-Leste 0 0 4 0 4

Haiti 0 1 2 0 3

Total 9 countries 3 5 31 7 46

Of which DAs 2 0 24 7 33

Of which TAs 1 5 7 0 13

A few elements have influenced the limited effect of DC on inter-sectoral division of labour. Firstly, there was a lack of a clear communication that DC should only be used in focal or non-focal sectors. Secondly, broad sector definitions offered the possibility to enter various sub-sectors. Thirdly, the existence of thematic facilities, designated non-focal sectors to which budget is allocated in the country strategy documents, and ad-hoc support led to projects which operate alongside the three focal sectors. In addition, these results should be seen in a context of little adherence in general, i.e. by all donors, to inter-sectoral division of labour on the basis of three focal sectors per donor. Use of DC in both focal and non-focal sectors The original communication on DC described the instrument as a means of implementing the ‘Code of Conduct on Division of Labour’, by using DC in non-focal sectors to become a passive donor. However, in only one of the programmes in the case studies (DA with KfW in Tanzania), DC has been used to exit a sector. In 20 of the 31 cases, the EU signed a DA for a programme in its focal sector, thereby by definition limiting an effect on inter-sectoral division of labour, and in three of the 13 cases it signed TAs to carry out a programme in an EU non-focal sector. However, also in cases where the EU signed a DA for a programme in a non-focal sector, this was rarely as part of an exit strategy. Reasons for signing a DA in a non-focal sector were related to emergency actions (a DA in Haiti after the earthquake), or to actually become more active in a particular sector (illustrated in the two cases with a negative effect in Mali). Therefore, even in cases where the DA was in an EU non-focal sector, it did not contribute to an improved inter-sectoral division of labour. Only in two cases, the EU assumed a more or less passive role – these are the only two DAs in Tanzania for which a strong effect was recorded. Broad sector definitions The other five negative scores are related to the broad sector definitions employed for some countries. In Nicaragua and Palestine, the DA contracts were part of a focal sector, but these sectors were so broadly defined that it led the EU to become active in a new (sub-) sector. Had the EU remained a silent donor, the effect would not have been negative; however, the EU joined the policy dialogue in this (sub-) sector, thereby increasing the number of donors. An additional explanatory factor for the absence of

15


positive scores in Palestine is the urge to be visible and to have ‘a seat at the table’. Donors, including the EU, are in that context more inclined to be active in many sectors. Additional lines of funding alongside the focal sectors The GCCA case study has illustrated the friction that a thematic programme can have with the focal sectors at the country level. The EU received under the GCCA seven TAs to fund 12 projects in climate change in specific countries, while climate change was not in every country part of an EU focal sector. For example, in Timor-Leste and Mozambique, environment and climate change were not focal sectors but the EUD had to implement the funds of the TAs which contributed to the GCCA. Although it delegated the funds to other partners via DAs, the EU became active in the sector. In addition, the cross-cutting areas defined in the National Indicative Programmes (NIPs) enabled the EU to fund projects and programme alongside its focal sectors. DC was however sometimes used effectively to deal with these funds and facilities additional to the focal sectors: in Tanzania, the EU had to spent money in the water sector because of the allocations from the MDG Initiative. It was decided to delegate those funds to KfW, in line with the Joint Assistance Strategy for Tanzania (JAST) at that time, which indicated that the EU would exit the sector. While the additional support to specific, cross-cutting themes such as climate change is desirable, it should be noted that these interventions operate outside the traditional division of labour on which the design for DC is based. In addition to these elements at the level of the EU, other donors also employ generic sector definitions and are often active in multiple sectors at the same time. For 26 out of the 30 DA-funded programmes, the sector was a focal sector of the DC partner21. While this is in line with inter-sectoral division of labour, it should be mentioned that few donors employ the three-focal-sector concept and often use broadly defined pillars or goals. That is further illustrated by the fact that 10 of the 13 TA-funded programmes could be attached to a priority sector of the TA partner. Context-specific issues are also at stake for the DC partner. The ‘seat at the table’ mentality in Palestine has resulted in donor proliferation in many sectors. The absence of mainstreamed donor strategies focusing only on a limited number of sectors in all case study countries, combined with a couple of cases (e.g. Nicaragua, Palestine) where sector definitions were very broad, made it difficult for DC to adhere to a formal division of labour strategy. Tanzania was the only case where the formalisation of division of labour was more advanced through the JAST. Pragmatic rather than strategic Division of Labour Although the inter-sectoral division of labour was hardly improved by DC, it was found that often the DC agreement contributed to a better intra-sectoral division of labour. When the three cases that were scored strongly on inter-sectoral division of labour and a case which consisted of a transfer of General Budget Support (GBS) are taken out of the equation, it was found that out of the remaining 42 DC-funded programmes, 34 had a positive effect on intra-sectoral division of labour. Within broadly defined sectors, DC was used to facilitate a division of labour between sub-sectors. DC was used in Palestine to foster a joint approach in the solid waste sub-sector; in Ghana, the use of DC was recommended on the basis of a mapping study, which suggested a contribution to two multi-donor pooled funds, which were led by the lead donors in the respective sectors. However, in most cases, the intra-sectoral arrangements were based on pragmatic considerations, meaning that the DC agreement was signed to delegate support to a DC partner with additional experience. In Mali, the EU was active in large-scale irrigation, but delegated assistance to small-scale to GIZ, which was active in

21

For three DAs it was not possible to distinguish what was the focal sector of the DC partner.

16


this specific sub-sector. In Tanzania, DFID was an active donor in the roads sector, but transferred funds to the EU for the construction of trunk roads. Findings from the questionnaire The results of the questionnaire support the finding that exiting a sector was not the main motivation behind a DC agreement. Of the 33 EUDs22, 28 (85%) responded that the motivation “DA(s) was/were used as an instrument to phase out the EU support to a non-focal sector” was “not really an issue”. On intra-sectoral division of labour, 24 respondents (72%) noted that the statement ‘DAs would improve the division among donors within a sector’ was a motivation for preparing and concluding a DA. While this is clearly a higher score than for inter-sectoral division of labour, only a quarter of the respondents saw division of labour as a main motivation; almost half of the respondents considered it a secondary motivation. The figures below indicate that the general perception on the effect of DC on both inter- and intra-sectoral division of labour is more positive than negative. However, the comments made reference to issues found during the case studies, such as the lack of adherence to focal sectors by other donors and the fact that DC was used in focal sectors and was more motivated by a comparative advantage of the DC partner than by a strategic decision. Figure 3.1.1 EUD questionnaire response: To what extent has Delegated Cooperation resulted in a better division of labour amongst donors across various sectors?

Response Total % of responses

0% 20% 40% 60% 80%

%

Strongly 5 14

Modestly 17 49

No change 9 26

Deterioration 1 3

Don't know 3 9

Total respondents: 35

Figure 3.1.2 EUD questionnaire response: To what extent has Delegated Cooperation resulted in a better division of labour amongst donors within the sector of the DC-funded project(s) or programme(s)?

Response Total % of responses

0% 20% 40% 60% 80%

%

Strongly 13 37

Modestly 13 37

No change 6 17

Deterioration 1 3

Don't know 2 6


22

Question was only posed to the EUDs with DAs (33).

17


Contribution of improved division of labour to reduced transaction costs The original DC guidance was based on the merits of inter-sectoral division of labour, which would reduce transaction costs. As DC has only contributed to an improvement in inter-sectoral division of labour to a very limited extent, this should not have an effect on the reduction of transaction costs. This is because the two donors remain active in the sector and engage in policy dialogue, and they are not silent partners. There is therefore no effect on the total workload. However, within the sectors, DC has been used to better organise the work, usually by delegating implementation to the donor with a clear comparative advantage. Some positive examples were found (e.g. in Timor-Leste) where the EU and DC partners agreed on a division of labour within the sector in which the partners focused on the implementation and the EU led the policy dialogue.

3.1.5 More co-financing (JC 1.2)

JC 1.2. Delegated Cooperation has led to more co-financing of projects and programmes, which has contributed to a reduction of transaction costs

Co-financing refers to the notion that an action (a programme, including a sector programme, or project) is funded by more than one actor (public or private donors). It has grown as an aid practice against the background of an increasingly fragmented aid landscape. In the early 2000s, co-financing became popular within the aid community as a tool to reduce transaction costs and to improve alignment to country systems and programmes. In that period so-called government-managed pooled (or basket) funding of sector programmes increased rapidly. More recently though, questions have been raised about the effectiveness of such co-financing arrangements, in particular in terms of their actual use of country systems23. From the perspective of the EU, DC was consistently associated with joint co-financing throughout the period 2007-2012. The first instruction note on DC, published in late 2007, was even titled: ‘guidance on joint co-financing with Member States and other bilateral donors’. Funding from two or more donors was formulated as a minimum condition for entering into a DC agreement. In subsequent instructions in 2008 and in a review report on DC in 2012, DC and co-financing were again presented as intertwined concepts. In the ‘Guidance paper on Delegated Cooperation with Member States’ issued in 2012, DC was explicitly presented as ‘a joint co-financing modality, by which one EU donor entrusts the management of funds to another’. The link between DC and co-financing was further confirmed by the statement that through DC “it is possible to concentrate the implementation of similar actions in one hand (the fund managing donor), thereby avoiding that several EU donors implement such actions in parallel. This should result in a reduction of aid fragmentation and proliferation, but also in a reduction of transaction costs due to economies of scale”24. An explicit link was therefore made between the pooling of resources through co-financing arrangements and reduction of transaction costs. Through co-financing, the programming, design, and supervision of the aid activities could then in principle be carried out by one agency on behalf of several donors. On the other hand, setting up a co-financing arrangement and coordinating the implementation of such an arrangement would create additional transaction costs. Nevertheless, it was expected that there would be a net positive effect on reducing transaction costs at the level of the donors. Transaction costs of the partner government were also expected to be reduced, because the partner

23

See for example Common Funds for sector support, Building Blocks or stumbling blocks, ODI Briefing Paper, February

2008. 24

Guidance paper on Delegated Cooperation”(2012).

18


government would not have to consult each donor bilaterally on various aspects of project design and implementation, when only one single fund management system could be used. When analysing co-financing, a distinction has to be made between parallel co-financing and joint co-financing. In the case of the latter, the contributions of the various donors are put together in one basket and there is only one single management system. Moreover the donor contributions are not earmarked for funding specific project expenditures. In the case of parallel co-financing, the donor contributions are usually earmarked, while there are specific donor related management systems (although some aspects may be joint). The DC guidance has referred in general to ‘joint’ co-financing and the cost savings mentioned in the previous paragraph have predominantly been associated with joint co-financing. Findings from the portfolio analysis In the desk phase, the documents for all DC contracts were studied to identify the nature of the funding. For the entire portfolio of 164 DAs, the findings are as follows:

78 DAs were part of a joint co-financing arrangement (48% of the total);

19 DAs were part of an arrangement consisting of both joint and parallel co-financing (12% of the total);

25 DAs were part of a parallel co-financing arrangement (15% of the total);

41 DAs were not part of a co-financing arrangement (25% of the total), meaning these DAs were linked to programmes fully financed by the EU and implemented by a DA partner;

Documentation of one DA did not provide sufficient information for making a classification.

Thus, about 75% of the DAs in the entire portfolio were part of a co-financing arrangement (joint and/or parallel). These figures on joint and parallel co-financing should however be interpreted with caution, because during the field visits it was often found that a co-financing arrangement presented as joint co-financing was in fact a form of parallel co-financing, with earmarking and separate administration of the funds of the co-financing partners. When looking at the trend over the years, there has been an average of approximately one-quarter of the DAs not being co-financed. The number of non-co-financed DAs has increased in 2013 and 2014 in absolute and relative terms. Table 3.1.2 Portfolio analysis: Number of DAs co-financed between 2008 and 2014

Year 2008 2009 2010 2011 2012 2013 2014 Total

No 1 2 0 6 4 12 16 41

Joint co-financing 2 6 9 19 9 25 27 97

Parallel co-financing 2 0 0 3 5 5 10 25

Total 5 8 9 28 18 42 53 16325

Not co-financed DAs as % of total 20% 25% 0% 21% 22% 29% 30% 25%

The analysis of the entire portfolio of 59 TAs has shown that co-financing is an integral part of the TAs. 54 of the 59 TAs (92%) were/are co-financed by the EU, and in most cases with significant amounts. Only five TAs linked to the GGCA programme were not co-financed by the EU; Four of them were financed entirely by the TA partner while in one case other donors than the EU provided co-funding. Additional details are presented below:

25

No information found for 1 DA.

19


20 TAs were co-financed by the EU and the TA partner;

11 TAs were co-financed by the EU and various TA partners;

1926 TAs were co-financed by the EU, the TA partner and other means of co-financing arrangements;

One TA was not co-financed by the EU, but by other donors.

There were five TAs where the TA partner was the only funder27.

No co-financing information was available for three TAs. Findings from the case studies In the following table, the scores regarding 46 DC projects and programmes are summarized, which shows that 22 out of the 46 projects (48%) were joint co-financed. Out of the remainder of the programmes, 14 (30%) were co-financed in parallel. In 10 cases (22%) the programme supported by one or more DC agreements was not co-financed at all. While some form of co-financing was therefore present in 78% of the case study programmes, the lack of other funding in 22% of the cases is unexpected given the emphasis put on co-financing in the DC guidance. Table 3.1.3 Case study scores: Effect of DC agreements on co-financing

Country Joint co-financing

Parallel co-financing

No co-financing

Total

Benin 2 4 1 7

Ghana 3 0 0 3

Mali 2 2 0 4

Mozambique 5 1 0 6

Nicaragua 0 0 4 4

Palestine 7 4 1 12

Tanzania 3 0 0 3

Timor-Leste 0 0 4 4

Haiti 0 3 0 3

Total 9 countries 22 14 10 46

Of which DAs 11 13 9 33

Of which TAs 11 1 1 13

Source: calculated on the basis of CRIS data retrieved in 2015.

In 18 cases, a strong effect on improving co-financing was found in the case study countries. The DC agreement stimulated co-financing by contributing a significant amount to a bilateral project or enlarged a multi-donor basket fund. In four cases, although there was joint co-financing, the contribution was modest. For two out of the 22 cases, a combination of joint and parallel co-financing was found. It should also be noted that sometimes the difference between joint and parallel was very small. As mentioned above, in some cases (e.g. in Haiti) the DC Action looked like a joint co-financing arrangement because one single management systems was used for all funds, but the funds of each donor were earmarked implying that it was in fact a case of parallel co-financing. It was observed that there were occasionally clear motivations to prefer parallel co-financing above joint co-financing. In the case of some projects in Benin, the EU was particularly interested in funding only a specific component as the DC partner had a particularly strong comparative advantage as regards that component. Also, in some

26

Including 13 TAs to PEGASE DFS. 27

The project funded by a TA of Ireland in Timor-Leste is counted as not co-financed, as the EU only covered

administrative costs.

20


specific cases in Palestine, the EU insisted on financing its own project components, e.g. in the TVET projects, but also in the KfW water project in Gaza, rather than contributing to a bigger programme. There were also cases where visibility of the EU was a reason to refrain from joint co-financing. The absence of co-financing turned out to be partly country-specific. While in the other countries, the majority of the programmes were co-financed, in Timor-Leste and Nicaragua, all DC-supported interventions were not. In Nicaragua, where budget support was suspended, DC provided a flexible and swift alternative implementation modality without creating additional workload for the EUD. In Timor-Leste, DC offered a solution to staff shortages within the EUD. Because of these strong motivations to use the DC modality, creating more co-financing was of lesser importance. When speaking about the absence of co-financing, it is worth to also consider the position of the DC partner. The co-financing principle within a DC context may not be equally shared among EU Member States. In the case of Belgium for example, their guidelines on Delegated Cooperation make no mention of co-financing and implicitly suggest that the partner should finance 100% of the action. It is also important to note that co-financing in the case of Belgium comes from the Belgian Ministry of Foreign Affairs, while the Belgian Technical Cooperation (BTC) is the implementing agency. The difference between the donor (i.e. the Ministry) that has to provide the funding and the DC partner, which is at times a different entity, has to be borne in mind. It should be noted that in some cases, the EU or the DC partner contributed to funding management or overhead costs. This was the case with all DAs implemented by GIZ. The latter indicated that the 7% management costs included in the DA budget, were not sufficient to cover GIZ’s management costs. Therefore GIZ received additional funding from the Ministry (BMZ) to cover the additional management costs. Since that funding was purely meant to cover part of the operational costs of GIZ itself, and did not add to funding project/programme activities, it has not been regarded as actual co-financing. The GCCA project in Timor-Leste is another example, where the funds from the TA concluded with Ireland were entirely used for the project, while the EU only added some funds to cover M&E and audit costs. Generally, it can be considered that a TA is always co-financed. Although five instances were found where the EU did not provide additional funding to a TA-funded project, this can be explained by the use of a TA in the context of a thematic facility (GCCA). Sweden and Ireland signed overarching TAs, transferring funds to the GCCA to be used for multiple projects. The GCCA management preferred to use the funds to enlarge the geographical scope of the facility, i.e. providing support in as many countries as possible, instead of funding large, co-funded programmes in only a few countries. As such, in a much broader sense, it could be said that the TAs provided co-financing to the overarching climate change initiative. In line with this finding, it was observed that the nature of TAs per se strongly encourages joint co-financing. According to the EU Financial Regulations, the funds transferred from another donor to the EU are treated as assigned revenues. This implies that they are included in the EU budget (specific commitments and payments appropriations) or in the EDF and their management is governed by EU procedures. On the other hand in the case of DAs, DA partners have to comply with the EU Financial Regulations, which do not encourage joint co-financing. Furthermore it was found that TA funds tend to be less earmarked than DA funds. In the majority of the cases, TAs were part of joint co-financing arrangements rather than parallel co-financing arrangements. Findings from the questionnaire The pooling of funds (joint co-financing) was considered by 13 of the 33 respondents to the EUD questionnaire as a main motivation for signing a DA, while 12 respondents

21


considered it not really being an issue. Five respondents perceived it as a secondary motivation. The remaining three respondents were of no opinion. While the majority of respondents therefore perceived joint co-financing as having been considered in the decision making, a relatively large minority did not think however that it was an important issue. Out of the 18 EUDs that dealt with TAs, 13 considered the pooling of funds as either the main or a secondary motivation. However, this question might be less relevant as TAs are decided at Headquarter (HQ) level. The figure below presents the responses from EUDs to the question whether DC has led to more joint co-financing in both sector and DC project/programme. Although this figure suggests that for 43% and 31% of the respondents DC has not contributed to more joint co-financing respectively in the sector or DC programme, it should be noted that a large number of respondents did not answer the question and from the comments made in response to the question, it appeared that the situation was often more complex. Figure 3.1.3 EUD questionnaire response: co-financing (total responses: 35)

Nevertheless, the responses made clear that the positive link between co-financing and DC is not systematically recognised. One comment clarified that it is not fully true that DC is encouraging co-financing by mentioning an example where DC gradually led to the DC partner pulling out of the sector. When asked about the effect of more co-financing on the reduction of transaction costs, four respondents indicated a strong effect, 19 a modest one and eight responded that it had no effect. Contribution of more co-financing of projects and programmes to reduced transaction costs In 22 of the 46 projects/programmes examined in detail during the field missions, joint co-financing was found. For these cases, the link between co-financing and reducing transaction costs is strongest: as noted above, the reduction of these costs mainly derive from savings in sharing management systems and refraining from earmarking. This is to a lesser extent the case for the parallel co-financed projects. However, for some cases of parallel co-financing, a management system was shared and funds were integrated. In 10 cases, co-financing was not found and has not contributed to reducing transaction costs.

3.1.6 Larger programmes (JC 1.3)

JC 1.3. Through Delegated Cooperation the programmes/projects became larger (budget and/or scope) which has led to a reduction of transaction costs

18

12

11

15

6

8

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

DC has led to more joint co-financing ofthe project(s) / programme(s) supported

by the DC arrangement

DC has led to more joint co-financing inthe sector(s) supported by the DC

arrangement

Yes No Don't know

22


This JC assumes a positive relationship between the size of DC projects and programmes and a reduction of transaction costs. The analysis focuses on the volume of DAs and TAs in relation to the total programme or project budget both at the level of the entire portfolio and at the level of the projects and programmes covered by the case studies. In the 2012 ‘Guidance paper on Delegated Cooperation with Member States’, the idea that DC would lead to larger programmes was explicitly mentioned in relation to co-financing (see JC1.2). The link between DC and co-financing was confirmed by the statement that through DC “it is possible to concentrate the implementation of similar actions in one hand (the fund managing donor), thereby avoiding that several EU donors implement such actions in parallel. This should result in a reduction of aid fragmentation and proliferation, but also in a reduction of transaction costs due to economies of scale”. It was also mentioned that DC was considered as a “modality to move towards larger programmes …. jointly funded with other public or private donors”. Within the literature on aid effectiveness, there is a general agreement that – independently of the nature and degree of alignment of a project – the balance between cost and benefit of Delegated Cooperation partly depends on the size of the project/programme. Although an accurate cost-benefit assessment cannot be made, in general, the probability that the benefits outweigh the costs is higher if the cooperation focuses on large projects or programmes28. Meanwhile, such cost-benefit analysis should also take into account the size of the actual EU contribution to the Delegated Agreement and not just the size of the project/programme it contributes to. The recent guidance (2012) issued by the Commission that the delegated EU amount should not be less than €3 million is likely to be based on that assumption. In its Report on Delegated Cooperation 2007-2012, the Commission noted that ‘’the current situation with various small Delegated Cooperation arrangements is not viable in view of the level of staff resources required at both Delegation and Head Quarters level’’. Findings from portfolio analysis The average budget of the 164 DAs amounted to €7.7 million, with 41 DAs (25% of the total) having a budget lower than the threshold of €3 million (established in 2012)29. The average budget size of newly approved DAs increased gradually from €2.8 million in 2009 to €10 million in 2013, but declined to €8.3 million in 2014 (see table below). Table 3.1.4 Portfolio analysis: Delegation agreements – breakdown per year

Year Number of DAs Value (mil. €) Average value No. of

agreements < € 3 million

2008 5 48.3 9.7 1

2009 8 22.4 2.8 4

2010 9 31.1 3.5 4

2011 29 150.1 5.2 10

2012 18 154.1 8.6 4

2013 42 419.7 10.0 10

2014 53 437.7 8.3 8

TOTAL 164 1,263.4 7.7 41 Source: calculated on the basis of CRIS data retrieved in 2015.

In 39 cases (24%), these DAs were the sole funding source of the project concerned, which means that the DA Action consisted of the entire project, and was not embedded in a larger programme. In 67 cases (41%) the project supported by the DA was also

28

Nordic Plus, Practical Guide to Delegated Cooperation, October 2006, pp. 13. 29

Seven of these small DAs where in Small Island Developing States.

23


supported by other funding sources (for example, the DA Action was complementary to a broader budget support programme), while in 16 cases the project concerned was supported by more than one DA. In 41 cases it was not possible to establish the composition of the funding of the project concerned. Some more precise figures could be retrieved from the CRIS database about the total budget of 89 projects supported by DAs30. The average budget size of these 89 projects amounted to €15.9 million, with 37 having a budget of less than €10 million. The DA contribution to the budgets of these projects and programmes represented 50% on average. Unfortunately no data were available allowing to calculate a trend from 2008 to 2014. On the basis of the DA contracts and Action Fiches only, it appeared to be relatively difficult to establish whether a DA project is part of a larger project or programme of another donor. Nevertheless, for the whole portfolio, for at least 20 DAs it was indicated that they were part of a larger programme of another donor. The trend in the average budget size of newly approved TAs was quire erratic (see table below), with peaks in 2010 and 2013, an average budget size of €4.9 million and 35 TAs (57% of the total) having a budget lower than the threshold of €3 million (established in 2012). Table 3.1.5 Portfolio analysis: Transfer agreements – Breakdown by year

Year Number of TAs Value (mil. €) Average value No. of agreements

< € 3 million

2008 4 5.8 1.5 4

2009 8 38.3 4.8 5

2010 9 84.2 9.4 4

2011 12 31.9 2.7 7

2012 9 30.4 3.3 5

2013 10 78.1 7.8 5

2014 5 20.3 4.1 3

No year specified

31

2 2.0 1.0 2

TOTAL 59 291.2 4.9 35 Source: calculated on the basis of CRIS data retrieved in 2015.

CRIS does not provide information on the budget size of all projects supported by TAs, but more information has been collected from documents related to the 40 TAs of the case studies. That information reveals that the average size of the projects supported by those TAs amounted to €200 million, while the average size of a TA was € 5.1 million. However, this comparison is distorted by the fact that the average size of the supported projects was strongly influenced by two large budget support programmes (in Ghana and Mozambique) and by the enormous size of the PEGASE instrument, to which 13 TAs contributed. Thus, the average size of the supported projects and programmes is relatively substantial, but the average size of the DAs (€ 7.7 million) and the TAs (€ 4.9 million) is much smaller, while there are still DAs and TAs with a budget below the threshold of € 3 million. Findings from case studies The table below shows the scores from the case studies regarding 46 projects and programmes supported by DAs and TAs. In 33 projects and programmes (72%) a positive

30

Total budget = DA plus funding from other sources. 31

Including 2 TAs from Luxembourg, of € 1 million each. These TAs were found during the field mission; it could not be

clarified in which year they were signed.

24


effect of DC on the programme size was measured, while for 20 programmes there was even a strong effect. In four cases there was a negative effect. Table 3.1.6 Case study scores: Effect of DC agreements on increasing the size of projects and programmes

Country Strong Effect

Modest Effect

No change Negative

effect Total

Benin 3 2 2 0 7

Ghana 0 2 1 0 3

Mali 3 1 0 0 4


Nicaragua 0 0 4 0 4


Tanzania 3 0 0 0 3


Haiti 3 0 0 0 3




The finding from the portfolio analysis that DC positively contributed to the size of projects and programmes is confirmed by the case study analysis. This is also in line with the findings on joint and parallel co-financing. The 10 programmes that had no co-financing and the one case where only a small amount of parallel co-financing was found scored either no change or a negative effect on increasing the size of a larger programme. In Mozambique and Ghana, two programmes scored high on joint co-financing, but this did not affect the size of the programme – as it merely meant existing funds were now bundled together. For the 13 TA programmes, the effect on the size of programme is positive for 11 programmes (85%), while for the 33 DA projects and programmes a positive effect was found in 22 cases (67%). This indicates that the effect is more positive in the case of TAs. However, this should be nuanced as the volume of the TA contribution compared to the total programme budget is rather limited. A good example is the PEGASE mechanism of Direct Financial Support (DFS) to the Palestinian Authority32. In some other cases (Palestine PEGASE private sector support with TA from Japan, Swedish TA to a Programme for Local Economic Development in Mozambique, and three TAs to a governance-public finance management programme in Benin) the use of the TA funds was not proportionate to the execution of the entire project budget (funded by the TAs and other sources). In those cases, the TA funds were used first before other sources of funding were used. On the other hand there was a project in Palestine and another one in Benin where not all EU funds were spent when the project was closed, indicating that TA had been larger than the funding need. Regarding the DA programmes with a strong effect on the size of the programme or project, some countries such as Haiti and Tanzania show only positive scores, while in

32

Three countries – Austria, Belgium and Luxembourg- concluded 12-13 TAs regarding the PEGASE DFS mechanism

with a total volume of € 29.2 million. The information is CRIS was incomplete and mentioned 8 TAs with a total volume

of € 16.2 million. In total, 19 countries including the EU contributed to PEGASE. The EU was responsible for more than

85% of all funding. The TAs formed only a minor part of the other contributions as most Member States made use of a

Memorandum of Understanding. The 12-13 TAs and the volume of € 29.2 million represented 9% of the total Member

States contributions.

25


Palestine most scores are also positive. In Haiti, one key explanatory factor (for the PARAQ project) was that the EU had to respond quickly to needs in the aftermath of the earthquake, including in non-focal sectors. Contributing to another donor’s programme through a DC was a pragmatic response. In Tanzania, the two DAs and one TA all contributed to substantially larger programmes, which can be explained by the relatively good donor coordination and harmonisation as reflected in the JAST. In the fragmented and politically sensitive Palestinian aid context, like-minded donors are interested to avoid overlap between projects and to create synergies where possible. DC is one instrument to do so if visibility and ‘a seat at the table’ are assured. Therefore, the EU aimed, through DAs and TAs, to support more comprehensive projects. There is a clear relation between co-financing and the size of projects and programmes. In those cases where negative effects of DC on the size of projects and programmes were reported, an important explanatory factor was that these were stand-alone programmes without co-financing and where the implementation was delegated to two DC partners rather than one. The GCCA case study offered an ambiguous picture. On the one hand, the TAs signed to contribute to the GCCA should be seen in the context of the larger climate change programme. On the other hand, the contributions were sometimes used to finance non-co-financed stand-alone interventions. Therefore, especially in the five cases where the TAs to the GCCA were not co-financed by another party, there was no effect towards a larger programme in the country. In other cases, the GCCA TA was used to promote sector budget support, so here the TA did contribute to a larger programme and prevented a separate intervention. Findings from the questionnaire Respondents to the questionnaire where asked how DC contributed to reduced transaction costs. 86% of the respondents (compared to 72% in the case studies) were of the opinion that DC leads to larger projects and programmes, which reduces transaction costs. This result is slightly more positive than the findings from the case study, which confirms that the perception of the DC output-outcome relation is positive.

Figure 3.1.4 EUD questionnaire response: Larger programmes in relation to transaction costs

Answer Total % responses

0% 20% 40% 60% 80%

%

Strong effect 16 46

Modest effect 14 40

No effect 4 11

Don’t know 1 3

Not applicable 0 0


More specifically, 71% of the respondents indicated that DC resulted in larger projects and programmes in terms of budget, while 63% indicated that DC contribute to larger projects and programmes in terms of scope and activities. When respondents were asked to rank the various effects of DC, it is interesting to see that 30% of the respondents (the highest score) ranked larger projects and programmes in the first place, and respectively 16% second and 23% third. The overall view that comes out from these answers is that DC’s

26


contribution to larger projects and programmes will reduce the transaction cost. However, the nature of that relation is not clearly specified. Contribution of larger projects and programmes to reduced transaction costs The findings from the desk review, case studies and the questionnaire point at a largely positive effect of DC on the size of projects and programmes. In principle, the evaluation should assess what would have happened in the absence of DA/TA, but often this counterfactual could not be established. Nevertheless, it is plausible that mainly for pragmatic reasons and to a lesser extent for strategic reasons, DC led to larger projects and programmes. The assumption is that the size of the programme had a positive effect on the transaction costs due to economies of scale (i.e. larger programmes translates to lower transaction costs per euro spent). This indeed might be the case, but whether this effect on transaction costs has materialised depends also on other factors such as the use of single management systems, the cooperation between DC partners (silent or active donor) and the degree of co-financing. Although it is plausible that there is a positive correlation between the size of programmes and reduced transaction costs, some case studies pointed at the risk of having too large programmes and projects, which might lead to counter-productive effects (dis-economies of scale).

3.1.7 Use of single management systems (JC 1.4)

JC 1.4. Through Delegated Cooperation at least two donors are using a single project/programme management systems, which has led to a reduction of transaction costs.

This JC assumes a positive relationship between the use of single management systems in DC projects and programmes and a reduction of transaction costs. Since the first guidance note on DC in December 2007, one of the five main aims is “encourage use of common monitoring, evaluation and accounting procedures”33 and this was repeated as one of the five strategic objectives of DC in the 2012 Guidance Paper on DC. The idea was that in this way the number of separate donor reviews and different accounting procedures could be reduced. This is captured in this evaluation under the output ‘single management systems’. As regards this JC, the issue is whether one single management system is being used - which can be the DC partner’s system, a hybrid system or a partner country system. The counterfactual is the management system(s) which would be in place without DC. Therefore, it is logical that DC will have a positive effect on the use of single management systems. The relation to co-financing is very strong. Improvement in single management system by definition requires co-financing, as it should bundle the funds of at least two donors under one system. In the case of joint co-financing, it is certain that at least two donors are making use of a single project/programme management system. In case of parallel co-financing, the relation is less strong, but not completely absent: the donor can opt for a parallel system, but could also use a joint system, and only earmark the funds for certain use. The output regarding single management systems (of donor funds) is also linked to the output on alignment, in particular the alignment with country systems (systems alignment). This will be explored under EQ5, in particular under JC 5.2.

33

Note from the Director-General of AIDCO to the attention of the Heads of Delegation, AIDCO D (2007) 24585, dd 04-

12.2007.

27


Findings from case studies The scores from the case studies are summarized in the table below. There appears to be a broadly positive effect of DC on the use of single management systems (27 of 46 programmes show a positive effect i.e. 59%, a majority of which with a strongly positive effect). For 10 projects and programmes there is no change (22%) and for 9 programmes and projects there is a negative effect (20%). The overall scores for DC vary more than the TA scores, which show a 77% positive score and only one negative score, reflecting the high level of (joint) co-financing in the case of TAs. Table 3.1.7 Case study scores: Effect of DC agreements on single management systems

Country Strong Effect

Modest effect

No change Negative

effect Total

Benin 4 1 1 1 7

Ghana 3 0 0 0 3

Mali 3 0 0 1 4


Nicaragua 0 0 0 4 4


Tanzania 2 1 0 0 3


Haiti 1 2 0 0 3




The results of the case studies indicate a positive trend. The management systems of the DA partner or the EUD (in case of a TA) were being used. As indicated above the issue of systems alignment is dealt with in EQ5. In most cases use was made of more or less standard project management arrangements, whereby a project management unit was created, partly staffed by external experts procured by the DA partner or the EUD (in case of a TA). Project Steering Committees were established in 35 of the 46 projects and programmes (76%; see EQ2 for more detailed information), mainly led by the partner country government. The use of single management systems is closely related to joint or parallel co-financing: where there is joint co-funding a larger programme is created that is managed by one partner. In these cases the scores showed a strong positive effect regarding the use of single management systems. In the case of parallel co-financing, project or programme components are sometimes still separately managed and therefore for some of these cases no change on this output is reported. All negative effects take place when multiple DAs are concluded for one and the same project or programme, which leads to a mixed or multiple management systems. If the EU itself would have implemented the entire project or programme, there would not have been a mixed or multiple management system. This appears to be a country-specific phenomenon and especially in Nicaragua and Timor-Leste, and to a lesser extent in Mali and Benin, this was the case. Findings from the questionnaire The responses to the questionnaire are in line with the findings from the case studies, though slightly more positive as 74% of the respondents (compared to 58% in the case studies) were of the opinion that DC contributes positively to reduced transaction costs through the use of single management systems.

28


Figure 3.1.5 EUD questionnaire response: Use of single management systems in relation to transaction costs


0% 20% 40% 60% 80%

%

Strong effect 13 37

Modest effect 13 37

No effect 5 14

Don’t know 2 6

Not applicable 2 6


In addition, 69% of the respondents were of the opinion that DC resulted in the increased use of a single management system of the DC-funded project(s) / programme(s). The responses to the questionnaire furthermore indicated that it was for most EUDs not a main motivation to save transaction costs by using common procedures. For the DAs, only 9% of the EUDs mentioned it as main motivation, and 39% as secondary, and for the TAs the percentages were respectively 17% and 28%. Also in terms of ranking of DC effects, the use of single management systems does not score very high. When respondents were asked to rank the various effects of DC, only 3% of the respondents ranked single management systems in the first place, and respectively 23% second and 16% third. This nuances to some extent the relatively positive score regarding the contribution of single management systems to reduced transaction costs; but could also reflect a limited understanding of the definition of a single management system. Contribution of single management systems to reduced transaction costs If joint co-financing is in place, DC is very likely to improve the use of single management systems. The 27 cases where the use of a single management system was promoted by DC probably led to a reduction in transaction costs as compared to a situation without DC. For the partner country, this is an absolute reduction in transaction costs. For the EU and the DC partner, this is a relative reduction – as it has to be offset against the costs of establishing a single management system (especially borne by the DC partner, e.g. completing the Pillar Assessment, and the fact that the DC partner systems should adhere to the EU Financial Regulations). These costs are further discussed under JC 1.6. When there was no co-financing, or parallel co-financing in which different systems were used to manage earmarked funds, often DC had little effect on moving to a single system and saving costs. In these cases, the project was implemented according to the regulations of the DC partner, but as no merging of multiple donor’s funds under one system took place, this had the same effect as when the donor would have used its own system. DC had a negative effect on single management systems when multiple DAs were combined within one programme. In this case, the acceptance of the DC partner’s system led to multiple systems within one programme, which resulted in additional transaction costs for the partner country, the DC partners and the EU.

29


3.1.8 Reduction of number of active donors (JC 1.5)

J.C. 1.5. Delegated Cooperation has reduced the number of active donors in the sector, which has contributed to a reduction of transaction costs

This JC assumes a positive relationship between fewer donors per sector and a reduction of transaction costs. The analysis focuses on the selected DAs and TAs in the sector context i.e. it is assessed whether DC led to a reduced number of donors per sector. Principle 5 of the EU Code of Conduct on Division of Labour states that EU donors, “with full participation and ownership of the partner country, will seek to limit the number of active donors to a maximum of 3-5 per sector, based on their comparative advantages. Other donors can still take part in sector activities by means of delegated cooperation modalities”. The last cited sentence implies that the transferring entity is considered to be a ‘passive’ donor. Therefore, DC can only reduce the number of donors per sector when a donor decides to delegate (or transfer) its sector support funds to another donor already active in the sector, and when that delegating donor stops being active in the sector (or will not become active in the sector otherwise). This JC is thus closely aligned with the improvement of inter-sectoral division of labour. Findings from the literature review Bürcky concluded that the number of donors per sector increased on average by 8.8% between 2005 and 2009 (therefore, an evolution in the opposite direction as envisaged by the EU Division of Labour principles, which aim at reducing the number of donors per sector)34. He noted also a difference between sectors, e.g. on average 13 donors in the sector Governance and Civil Society and seven in the Environment sector (averages of various countries in 2009). The number of active donors per sector is therefore country-specific and sector-specific as the interest in sectors to be supported changes over time. The sector definition also varies substantially with sometimes very broadly defined sectors with various sub-sectors or more narrowly defined sectors. In addition, some so-called sectors are defined in terms of cross-cutting subjects being relevant for (almost) all other sectors. A recent paper written by Nunnenkamp confirms the difficulty to aim for a reduced number of donors per sector35. Donors are ultimately accountable to their tax payers, and therefore aim to be visible in sectors and countries having been selected on the basis of their development cooperation policy, disregarding Division of Labour considerations. Political and/or commercial motives to provide support to a certain sector could also play a role, weakening the readiness of donors to join efforts for Division of Labour and reducing the number of donors per sector. Other studies aimed to determine what the optimum number of donors in a sector should be, but this proved to be very difficult36. This JC, however, assesses the contribution of DC to the number of active donors in a sector and is not directly concerned with the optimum number of donors per sector. Nevertheless, the findings have to be situated in

34

Urs Bürcky, Trends in In-country Aid Fragmentation and Donor Proliferation: An Analysis of Changes in Aid Allocation

Patterns between 2005 and 2009, Report on behalf of the OECD Task Team on Division of Labour and

Complementarity, 10 June 2011. 35

Peter Nunnenkamp, Albena Sotirova, Rainer Thiele, Do Aid Donors Specialize and Coordinate within Recipient

Countries? The Case of Malawi, AidData Working Paper 10, June 2015. 36


study, November 14, 2010. They conclude that the optimum lies somewhere in between total concentration (one donor

per sector) and extreme fragmentation. The optimum could probably best be described in qualitative terms as ‘big

enough to ensure a minimum level of competition and fruitful interaction between DPs, but small enough to keep

transaction costs low for both the partner country government and the DPs. Probably, the ‘3-5 donors per sector’ rule of

the Code of Conduct on Division of Labour is based on this logic.

30


the appropriate country and sector context in order to be interpreted with care regarding the contribution of this output to the overall outcome of reduced transaction cost. Findings from case studies In the following table the scores regarding 46 DC projects and programmes are summarized, which shows there is no change in the number of donors per sector as a result of DC in 28 out of 46 projects and programmes (61%). In addition, seven projects showed a negative effect (15%) and 11 showed a strong or modest positive effect (24%). A main finding is therefore that DC did not have an important effect on reducing the number of donors. Table 3.1.8 Case study scores: Effect of DC agreements on reducing the number of donors active per sector


Modest effect

No change Negative

effect Total

Benin 0 1 6 0 7

Ghana* 0 0 3 0 3

Mali 1 0 2 1 4

Mozambique* 2 2 2 0 6

Nicaragua 0 2 2 0 4


Tanzania 0 2 1 0 3


Haiti 0 1 0 2 3




There are several explanations why the effects of DC on the reduction of the number of active donors are limited. Firstly, this result is in line with the finding that DC was not used to exit a sector, which was one of the assumptions in the policy framework. Secondly, in the case of DAs the EU almost never became a silent partner, but remained active in the sector. This should not be considered necessarily as a negative effect, because in some cases the DC did lead to a good division of responsibilities between the EU and the DA partner as it allowed the EUD to become more active in the policy dialogue (e.g. Timor-Leste). Also in the case of TAs, TA partners remained active in policy meetings (e.g. in the PEGASE programme in Palestine and in Mozambique, where BTC contracted a technical assistant to be involved in the budget support policy dialogue). Thirdly, the effect of DC on the number of active donors was negative when there were more than one DAs per project (Timor-Leste), when the DC partner was not previously active in the sector (Palestine) and where the EU had to step after a natural disaster (earthquake in Haiti). Positive effects occurred where the EU decided not to interfere in new activities but left it to the DA partner (DA and TA in Mozambique to support ESPS II). Fourthly, the country context can be an important explanatory factor. For example, in Palestine there are a large number of donors active per sector and all donors prefer to have a ‘seat at the table’ in this politically sensitive context. Therefore, donors are not inclined to be a silent donor, but DC is an interesting instrument for them to reduce the workload to some extent and delegate implementation matters. Reduction of donors is also less applicable when DC is used in the context of thematic facilities, which are focused on increasing support for a certain sector. The motivation behind the GCCA is to generate more funding for climate change. In general, this output ‘reducing the number of active donors’ does not apply to the GCCA, as the aim of the initiative was to mobilise more (financial) support for climate change.

31


This explains why the EU did not seem to have been very concerned about the increase of active donors in the climate change sector in Timor-Leste. Findings from the questionnaire These main findings from the case studies are mainly in line with the responses to the questionnaire as indicated in the table below. 38% the respondents were of the opinion that DC contributed to a reduction of transaction costs because the number of donors was reduced (compared to 24% of the case studies). This result should be seen in the context of a general notion observed in the literature (see above) but also during interviews for this evaluation, that reducing the number is not by definition desirable. Figure 3.1.6 EUD questionnaire response: Reduced number of donors active in the sector in relation to transaction costs


0% 20% 40% 60% 80%

%

Strong effect 3 9

Modest effect 10 29

No effect 17 49

Don't know 3 9

Not applicable 2 6


In addition, 43% of the respondents did not see a reduced number of active donors in the sectors concerned as a result of DC, compared to only 17% that saw a positive result (with 31% no change, and 9% don’t know). Also in terms of ranking of DC effects, a reduced number of active donors did not score very high in the questionnaire. When respondents were asked to rank the various effects of DC, it is interesting to see that none ranked reduced number of donors in the first, second or third place. Contribution of reduction of number of donors to reduced transaction costs Given the fact that DC did not lead to a reduction of the number of donors per sector and that in most cases the DC partners remained active and did not become silent partners, the overall effect of this output on transaction costs is limited. There are some clear positive cases where this output positively affected transaction costs, but also some clear negative examples, while in more than 50% of the cases no effect was visible in the case study countries.

3.1.9 Limited additional transaction costs (JC 1.6)

J.C. 1.6. Delegated Cooperation has generated limited additional transaction costs

While DC is expected to reduce transaction costs through the outputs analysed under the JCs 1.1 to 1.5, it could also be that DC generates additional transaction costs in some areas both at the level of the EU and the DC partners. For example, drafting DA and TA Assessment Fiches and adjusting the DA and TA Action Fiches constitute additional transaction costs at the level of the EU, while both the EU and the DC partner might be confronted with additional work and costs when carrying out the six pillar assessments and bilateral negotiations when preparing the agreements. In some cases in Palestine, it was also indicated that the management of the DA itself took up quite a lot of resources.

32


If the EU is actively involved, the communication with the DC partner can generate an additional transaction cost. Findings from desk review In 2012 the EU reported in its report on DC the following on transaction costs and workload: “In particular, negotiations with delegated entities in the run-up to signing delegation agreements and the need to foresee the modality in Annual Action Programmes and in Financing Agreements have been seen as a burden. Monitoring efforts and the need to deal with derogations (e.g. N+1 or D+3 rules) have also contributed to this trend. As the workload rises for delegated entities when managing cooperation programmes on behalf of others, it also rises for EU Delegations, in the case of transfer agreements. Based on inputs received from the services involved, it is not possible to conclude whether the overall workload/transaction cost balance of all involved players in delegated cooperation has been positive or negative so far……” The interviews with the Headquarters of some DC partners conducted have also revealed concerns about the additional transaction costs. The following concerns have been expressed:

The process of getting a DA approved is seen as cumbersome by the DC partners interviewed so far, including the negotiations about the Financing Agreements, the time needed before having an EDF Committee approval and the lack of clarity about the EU procedures;

In theory, getting a TA approved should be less complicated than getting a DA approved, while also the procedures for concluding a TA are shorter and less cumbersome. In practice, in some specific cases, there are indications that TA negotiations can also be very lengthy, but in general TA General Conditions, i.e. the obligations of the EU towards the TA partner, are lighter than DA procedures;

The costs of the six pillar assessments are considered as very high and not entirely justified. For instance, some implementing agencies are questioning why they should go through this very thorough process, while they already have a triple A status. Moreover, when the new Financial Regulations entered into force in 2014, some DA partners had to undergo a new pillar assessment (7 pillar assessment);

Frequent changes of the EU procedures, templates, and Financial Regulations were said to further complicate and delay the various processes.

The management fees paid by the EU to the DA partner or received by the EU from TA partners were also mentioned as they are not a good indication of the actual (additional) transaction costs, because the latter may be higher or lower than the official fee expressed as a percentage of (part of) the contract amount:

For 159 DAs of the 164 DAs, the management fee could be retrieved37. The average management fee is 5.92%. For 101 DAs, the maximum fee of 7% was used;

For 34 of the 59 TAs, the management fee could be retrieved38. The average here was 3.3%. For 22 TAs, the maximum management fee of 4% was included – in three cases, the actual fee was even higher (up to 8%) as the TA did not include a percentage, but a fixed amount of the total project that would be used for management.

Findings from case studies The additional transaction costs related to the case studies could not be quantified. Nevertheless, the qualitative findings from the desk review were confirmed to a large extent during the case studies.

37

In five cases, the DA contract was missing or the management fee was not clearly mentioned. 38

This was lower than for DAs as TA contracts are not consistently saved in CRIS, and if the contract was available, it did

not always specify the management costs.

33


The following issues were raised:

The maximum administrative costs of 7% are according to some DC partners not sufficient and are not offset by cost savings due to efficiency gains. However, this depends to an important extent on the working methods and procedures of the DC partner of which some have a more labour-intensive way of working, which leads to higher administration costs. For example, in the case of GIZ the higher management costs are covered by BMZ even if the project is not co-funded by BMZ;

EU procedures remain very cumbersome and DC templates change too frequently, while still leaving a lot of room for differences in interpretation. This issue is mentioned both by EUDs and by DC partners;

Although the implementation of a DA project is expected to lead to a lower workload for the EUD, this is not always the case in practice. This is partly the case because the EUD wants to remain actively involved in monitoring and oversight even if the DC management system is being applied. The EUD rarely becomes a silent partner and remains active in the policy dialogue, which affects the workload. It should be realized that this may contribute to a better division of responsibilities based on complementarity and this is as such not always negative. In addition, a common feature of DAs is that the management load has been partly shifted to the DA partner being the agency implementing the project, and that the related costs are covered by the project budget. This is theoretically (in view of the definition of transaction costs) a saving of total transaction costs, but those ‘transferred’ costs could be called ‘hidden transaction costs’ covered by the project budget;

The involvement of TA partners in programmes and projects implemented by the EU varies to an important extent from relatively active to almost absent. In general, most TA partners experienced a reduction of their management costs as they are not directly responsible for implementation. In principle, the TA partners in the Palestinian PEGASE support mechanisms did participate in donor meetings and are regularly informed just as the Member States which contribute via Memoranda of Understanding (MoUs). Only TA partners that do not have a representation in the partner country are less active. This is a common phenomenon for the GCCA TAs as the money is regularly spent on projects in countries where the TA partners do not have an office. Some TA partners have special arrangements to remain actively involved as in the case of the Belgian GBS TA in Mozambique, where according to Belgian guidelines BTC has hired an additional Technical Assistant to participate in the policy dialogue. There are also cases (e.g. PAOSC in Mali) where the TA partners said they saved costs associated to managing programmes with non-governmental organisations and used those savings to be more active in the policy dialogue;

While the EU commits to using the systems of the DC partner, the implementation of the Action still has to be in line with the EU Financing Regulations (FR). The report on DC of 2012 already raised the issue of the difference between ‘sub-delegation’ and ‘sub-contracting’. In cases where the DC partner delegates the responsibility to procure and contract to a third party, lengthy negotiations and eventually changes in the DA General Conditions were needed to reflect the changes in liability. Eventually another pillar ‘sub-delegation’ to the pillar assessment was added, that had to be completed by the agencies that wanted to sub-delegate parts of the budget implementation tasks to another entity. Another example is the adherence to the D+3 rule, which entails that the fund managing entity has to complete all procurement and contracting within three years after the signing of the Financing Agreement. Because the Delegation Agreement is sometimes signed months after the FA, this leaves little time for the DC partner to use all the funds. To summarize, these rules prevent the DC partner from completely using its own framework and can result in additional transaction costs. One respondent to the questionnaire noted: “There was a negative factor in that the Delegation Agreement was a hybrid one, with

34


the DC partner having to apply EDF procurement rules. This undid a lot of the advantages of using the organization's own rules, and caused a lot of difficulties.”39

Findings from the questionnaire The responses to the questionnaire indicate that reduction of the EUD workload was considered in two thirds of the responses as a main or secondary consideration for concluding a DA. Figure 3.1.7 EUD questionnaire response: The DA modality would reduce the EUD's workload


0% 20% 40% 60% 80%

%

Main motivation 6 18

Secondary consideration 16 48

Not really an issue 11 33

Don't know 0 0


39

Responses to EUD questionnaire, Annex 5, p.65.

35


3.2 EQ 2: Ownership and leadership

EQ 2: To what extent has DC strengthened the ownership and leadership of the partner countries as regards the DC funded projects and programmes and the policy formulation and implementation in the sector of the DC project or programme?

3.2.1 Definition and scope

Ownership” is defined by the OECD-DAC as ‘’the effective exercise of a government’s authority over development policies and activities”40. Following this definition, EQ2 focuses on assessing the ‘effective exercise of the partner government’s authority over the DC project or programme and the sector development strategies it is part of. ‘Effective exercise’ has been measured by assessing the involvement of the partner country government in the formulation, implementation and monitoring of the DC project or programme and the relevant sector strategy (see JCs 2.1, 2.2 and 2.3). Subsequently, the extent to which DC has possibly strengthened the ‘effective exercise’ has been analysed. Special attention has been paid to analysing to what extent DC has contributed to a reduction of donors per sector and/or per DC project or programme, which in turn could have contributed to strengthening the ownership and leadership of the partner country government (see JC 2.4). Findings from the literature review Partner country ownership is recognized by donors as an essential element of improving aid effectiveness. However, strengthening country ownership has proven to be a cumbersome process in practice. The Working Group on Aid Effectiveness cites several studies conducted between 2006 and 2011, which identify a continuing heavy influence of donors on development policy choices of many partner country governments since the Paris Declaration has been signed41. The trend towards results-based management and ‘value-for-money’ of aid has, according to a number of studies, increased the pressure on partner country governments. This issue is also mentioned in a discussion paper published by the German Development Institute in 201442. Other studies indicate the lack of political will in some partner countries to strengthen their ownership and leadership of the development process. David Booth (ODI, 2011) argues that country ownership is often assumed, while it does not really exist. The paper calls for a long-term, sophisticated approach to establish country ownership by investing in leadership and institutions43. In the first High-Level meeting taking stock exercise of the progress made since the Busan High Level Summit of 2011, the need for increased country ownership was cited as one of the main challenges44.

40

OECD/DAC, 2006, Guidelines and Reference Series. 41

OECD, Strengthening ownership and accountability, 2011. 42

Sarah Holzapfel, Boosting or Hindering Aid Effectiveness? An assessment of systems for measuring agency results,

German Development Institute, Discussion Paper 31/2014, 2014. 43

David Booth, Aid effectiveness: bringing country ownership (and politics) back in, Overseas Development Institute,

Working Paper 336, 2011. 44

First High-Level Meeting of the Global Partnership for Effective Development Co-Operation, Mexico city, 15-16 April

2014.

36



Strengthened ownership and leadership of the partner countries has been identified as an envisaged outcome of DC in various DC documents. In an internal Commission note circulated in 2007, it was stated that “Delegated Cooperation should be considered in order to promote ownership and leadership of development programmes by partner countries and their accountability to people”45. The revised DC guidance paper (2012) mentions the same issue as one of the five strategic objectives of DC46. However, neither the initial DC guidelines (2007) nor the revised version of 2012 explain how the (assumed) causal relationship between DC and strengthened ownership and leadership will operate. It was indicated in the documentation that DC should be used in relation to budget support operations – maybe this expectation has led to the assumption that DC would strengthen ownership. Only one of the envisaged DC outputs of the IED has a clear link with this outcome, namely the reduction of the number of active donors in the sector (see reconstructed IED, figure 1.2 in Volume 1). The assumption on which this expected link is based is that with fewer donors, the partner country could have greater control over determining sector strategies and coordinating donor projects and programmes. However, this assumption should be treated with caution as much will depend on other factors such as the mode of operation and attitude of the donors involved (see also JC 2.4). Although this is the only output-outcome relation identified in the reconstructed IED, the Evaluation Team has taken a broader approach to assess the ownership of the partner country government, by analysing also the level of government involvement at different stages of the project/programme cycle (i.e. formulation, implementation and monitoring) and assessing what DCs have contributed to strengthening the government’s ownership and leadership.

3.2.3 Methodological observations

The OECD-DAC definition of ownership is in line with the definition presented in the Paris Declaration which reads: “partner countries exercise effective leadership over their development policies and strategies and coordinate the development actions”47. This definition is further elaborated by describing the commitments of the partner countries and the donors. Partner countries should (i) establish clear development policies and strategies; (ii) translate these strategies into prioritised results-oriented operational programmes expressed in medium-term expenditure frameworks and annual budgets, and (iii) take the lead co-ordinating aid at all levels. Donors should respect the partner country’s leadership and help strengthen their capacity to exercise it48. The Code of Conduct on Division of Labour (2007) underlines that “EU initiatives on a better division of labour will aim at reinforcing the objective of strengthening the partner country ownership and capacities to take over responsibility for donor coordination processes” and that reducing the number of donors in a sector could contribute to strengthening such ownership49. However, the policy and strategy documents of the DC instrument do not explain the cause-effect links between reduced number of donors and strengthened ownership and leadership.

45

Note from the Director General of AIDCO to the attention of the Heads of Delegation, AIDCO D(2007) 24585,

December 4, 2007. 46

Guidance paper on Delegated Cooperation, 2012. 47

The Paris Declaration on Aid Effectiveness, 2005. 48

The Paris Declaration on Aid Effectiveness, 2005. 49

EU Code of Conduct on Complementarity and Division of Labour in Development Policy, 2007.

37


For practical reasons, JC 2.2, dealing with implementation, and JC 2.3, covering the monitoring of the implementation, have been merged in this section. The findings in the case studies made clear that these two JCs are closely interlinked and that some of the related tasks carried out by the same institution. In the case studies a more in-depth analysis of ownership and leadership was made which went beyond what is written in official documents and reports. Views and perceptions were collected of stakeholders involved in the DC projects and programmes and in the related sector policies.

3.2.4 Formulation of policies and strategies (JC 2.1)

JC 2.1. DC has strengthened the partner country’s leadership as regards formulation of the policies and strategies of the sector of which the DC project or programme is part of, and of the DC project or programme itself.

Findings from the literature review Already in the early 2000s, it was emphasised that a sector strategy formulated by the partner country itself encourages country ownership of that strategy. In such cases, the government is more inclined to execute its self-formulated sector strategy, to plan the sector expenditures according to that strategy and to incorporate that expenditure plan into the overall budgetary and macroeconomic framework, than when those strategies would have been formulated by other persons or entities (e.g. the donors). Moreover, when such a country-owned strategy exists, it can serve as a guideline for donors to design and plan their support programmes50. The involvement of the partner government in the design of the DC project itself is equally important, because in principle the government is best placed to determine whether a project fits into the wider sector and national strategies and to rule out overlaps of various donor-supported programmes51. Without such involvement of the partner country, the DC project risks being or becoming irrelevant to the strategic priorities of the partner country. Findings from the case studies In the field phase, nine country case studies and one special GCCA case study has been carried out. The country case studies have covered 44 DAs and 31 TAs which are part of 46 projects and programmes while the GCCA case study dealt with three TAs of the country studies and nine additional TAs. Embedded Sector strategies The case studies have demonstrated that most of the sectors concerned have a sector strategy and that the DC agreements are based on these strategies (see also EQ5). In theory these sector strategies and policies are fully government-owned. In practice, however, some strategies are ‘home-grown’, while others are merely the product of external technical assistance and donor influence. The latter strategies are made by international advisers – not entirely “owned” by the Government - and often only partially implemented. Many of the DC agreements covered in the case studies are part of broader initiatives, which are often not sector specific. In those cases, the question about ownership of a sector strategy is less relevant: for example the four TAs being part of a Civil Society Support Project in Mali.

50

World Bank, World Development Report 2000/2001: Attacking Poverty, Oxford University Press, 2001, page 197. 51

DC agreements could refer to specific projects or broader (sector) programmes. For ease of writing often the term ‘DC

project’ is used, which could refer to either a specific project or a broader programme.

38


Furthermore, for several cross-cutting subjects, like environment & climate change, decentralization, civil society, TVET and health/HIV, a clear sector strategy does often not exist. In most cases these subjects or sectors are, to a limited or larger extent, covered in a broader national strategy or development plan. In the case of Mali, for instance, a decentralisation strategy was lacking to provide guidance to the DC supported projects/programmes. Two projects in Mozambique and Palestine, both funded by a DA, supported the design of a sector strategy in the HIV sector the TVET sector, respectively. Moreover, two other TAs are contributions to a General Budget Support programme (in respectively Ghana and Mozambique) which are meant to support the implementation of national poverty reduction strategies instead of being focussed on certain sectors. In both the GBS programmes, the relevant sector strategy is the national Poverty Reduction Strategy. One programme in Haiti is providing emergency assistance and is therefore also not directly aligned with a sector strategy. Another programme in Haiti has been in the transport sector where a national transport strategy was lacking. In Haiti, the lack of strategy development, the weak ownership and limited institutional capacity of the partner Government can be seen as challenges for successful project and programme implementation. The 12 TAs being part of the GCCA have been supporting partner governments to design suitable climate change strategies (for various sectors), the costing of these strategies and the design of an implementing framework. This suggests that these TAs do not directly improve country leadership; but they could support country ownership in an indirect way, by providing assistance in formulating climate change strategies and by supporting institutions and civil society organisations. Government involvement in project formulation Most of the documents related to DC agreements (regarding the 46 projects and programmes analysed) suggest some form of involvement of the partner country in the identification and formulation of the project and in the decision making regarding the DC agreement. The level of involvement has varied from one case to another and from one country to another. In most cases the Government has been consulted and it has endorsed the delegation of funds from one donor to another. The interviews with the EUD, the donors and also with the Government in the case study countries confirm this picture of consultation and endorsement. Another piece of evidence of ownership of the formulated DA project or programme is the fact that the EU has always signed a Financing Agreement with the partner country government agreeing on the DA arrangement as regards the EU support to be provided for that particular programme/project. Although signing a Financing Agreement is normal practice for aid provided by the EU, it is worth mentioning here, because it could have been expected that when the EU delegates funds and becomes a silent partner, it would let the DA partner sign an agreement with the partner country government. Also, the management mode (which is ‘indirect centralised management’ in the case of DC) needs to be mentioned in the agreement. Concluding a Financing Agreement with the partner country government regarding a DA demonstrates that the partner country government is at least aware of DC and has to ‘approve’. The Support to Ministry of Interior in Mozambique is an example of strong government ownership. The Government officially requested the EU to support its security sector reform. This request was the result of high level dialogue between the Government and the EU in 2007. The shared identification process started in March 2008 and the formulation started in one year later and was concluded in March 2010. In Gaza, in Palestine, the ownership of the Palestinian Authority has been relatively weak. The donors were funding projects, while the Authority had limited influence in Gaza and was only marginally involved in the formulation of the projects.

39


Findings from the questionnaire The questionnaire confirms the findings that the partner country has been involved in the decision making of the DC agreements. A small majority of 52% of respondents agrees with the statement while 40% disagrees. With regard to the results from the questionnaire, it needs to be mentioned that there might be a positive bias in the feedback from the EUD representatives, given their close involvement in the DC agreements. Figure 3.2.1 EUD questionnaire response: The partner country was involved in the choice / decision to use the DC modality


0% 20% 40% 60% 80%

%

Strongly agree 3 9

Agree 15 43

Disagree 10 29

Strongly disagree 4 11

No opinion 3 9


The response to the questionnaire was more convincing regarding the involvement of the partner country in the formulation of the DC project. This statement was supported by 76% of the respondents. Figure 3.2.2 EUD questionnaire response: The partner country was involved in the formulation of DC project(s) / programme(s)


0% 20% 40% 60% 80%

%

Strongly agree 3 9

Agree 24 69

Disagree 5 14


No opinion 1 3


3.2.5 Implementation & monitoring of policies (JC 2.2 & 2.3)

JC 2.2. DC has strengthened the partner country’s leadership as regards implementation of the relevant sector policies and strategies in general and the DC project/programme in particular

JC 2.3. DC has strengthened the partner country’s leadership as regards monitoring the implementation of the relevant sector strategy and of the DC project or programme in particular.

The assessment of these two JCs is carried out together because the findings for both JCs overlap to a large extent.

40


These JCs focus on establishing to what extent the partner government is leading the implementation and monitoring of the relevant sector strategies and to what extent the DC agreement has strengthened this leading role of the partner government. Leading the implementation of a strategy implies, among others, that the government verifies and assures that the objectives and strategies of individual projects and programmes are aligned with the approved sector strategies. The Government should also (be able to) prioritise projects and programmes and assure that donors are not all focussing their support on the same areas and/or subjects while ignoring others, and should be able to influence the scope and direction of the DC project during implementation. The level of influence of the Government at project and programme level is dependent on a number of factors. The Government might have a dominant role in the project implementation, for instance, by taking the lead in project management and/or it might participate actively in the Steering Committees (SCs), the main oversight body of many projects or programmes. In most donor-funded projects, an SC has been established to deal with strategic oversight and the direction of the project/programme. These SCs are also expected to play a crucial role in monitoring the implementation of the project, among others by reviewing annual implementation reports and approving annual work plans and budgets. The assessment of the monitoring of the DC projects and programmes pays attention to a leading role of the Government in the monitoring process in general and the contribution of the DC project or programme to strengthening the Government’s role in monitoring. At the level of the DC project itself, these JCs focus on establishing what the role of the partner government is in implementing and monitoring the project. In some cases the DC project might be implemented by a governmental institution, but in other cases by a semi-governmental or non-governmental organisation or by a Project Implementation Unit, possibly being controlled by the donors. The degree of involvement of the government in the management, coordination and monitoring of the DC project determines to a large extent the government’s capacities to influence the implementation strategies and to assure its alignment with the sector strategies. In addition, involvement of government bodies in the implementation of the DC project strengthens the government’s commitment to the project and most likely the sustainability of the results. It should be noted that these JCs are closely related with JC 5.2, which assesses the involvement of the government in management and steering of DC projects and the use of country systems. Findings from the case studies There are important variations in the involvement of the partner Government across the DC projects and programmes which have implications for the level of ownership and leadership of the government. The case studies have shown that these variations are to a large extent determined by the working methods and procedures of the DC partners and how they relate to the partner Government. (The distinction between the different types of DC partners is provided in chapter 1). This same issue comes back again in JC 5.2, related to the system alignment and the use of country systems. On the one hand, DC partners such as KfW and AFD tend to enter into a Financing Agreement with the partner Government (‘sub-delegation’). Subsequently, the Government usually signs an agreement with a national government body, often a ministry. Through this body, the partner country has a dominant role in the implementation of the project. The DC partner has limited influence on the policies or how the money is spent, but might apply strict monitoring. Thus, projects management by these type of agencies tend to make more use of partner country systems, for example the government’s own reporting templates. On the other hand, agencies such as GIZ, BTC, CICL and AECID are more directly involved in implementation themselves. Although there are differences in their way of operation, they mostly have their own implementation capacity to directly execute (some

41


of) the activities. For the other activities, they might sign contracts with service contractors, and involve the Government mainly through SCs. The latter ensures some role for the Government at strategic and/or monitoring level. Thus, these agencies tend to use their own systems, parallel to the government’s. They work with service contractors and make limited use of national government systems. For CICL the use of twinning arrangement might further strengthen the ownership of the Government on its projects. In general, this second type of DC partners are using mostly their own procedures, while the partner governments are less involved in project implementation than in case of direct management by the EU, in which the National Authorising Officer is typically more involved. Nevertheless, this limited beneficiary involvement and ownership needs to be qualified, because agencies like GIZ and BTC are also working on the basis of a participatory approach, involving the beneficiary institutions closely in the formulation, implementation and monitoring of the specific project activities. With regard to this, BTC has distinguished two different implementation modalities to manage and implement its projects and programmes. First, projects can be implemented ‘en régie’ (direct management) where BTC is signing all contracts and making all payments directly to the contractor following their own procurement rules (mostly for service and supply procurements). Yet, BTC can also implement projects under "co-gestion" (joint management) with a ministry or public body concerned. In this second way, BTC is co-signing all contracts and acting as contracting authority together with the ministry or public body concerned. This leads to a much stronger involvement of the partner Government. For DAs, BTC used its direct management modality in Mali and Tanzania. This was not by choice but reflected the fact that BTC had not been subject to the EU’s seventh pillar assessment. Outside the DAs, BTC commonly applied the more country-system friendly joint management modality in both Mali and Tanzania. The analysis of the case studies took the differences between DC partners into consideration and assessed the extent to which the projects with a DC agreement did have a SC. It revealed that most of the projects have established a SC at project or programme level. In 65% of the cases (30 out of the 46 of the projects and programmes) the project or project component where the DC intervention took place, has its own SC. In an additional five cases (11%), the SC has not been established at project but at a higher programme level. In some cases separate SCs exist at both levels. Table 3.2.1 Case study scores: Steering Committee exists in case study project and programmes

Country SC at project

level SC at higher

level no SC Total

Benin 5 1 1 7

Ghana 3 0 0 3

Mali 3 0 1 4

Mozambique 4 0 2 6

Nicaragua 4 0 0 4

Palestine 5 0 7 12

Tanzania 2 1 0 3

Timor-Leste 1 3 0 4

Haiti 3 0 0 3


of which DAs 24 2 7 33

of which TAs 6 3 4 13

42


In all cases with a SC, except one (the Support to Civil Society Pooled Fund, STAR in Ghana), at least one ministry, department or another government agency was member of the SC on behalf of the Government and in more than 90% of the cases in a role as chairman. In most DC agreement, the implementing donor is a full member of the SC and the delegating partner has joined as observer. For most projects and programmes, the SC meets (at least) twice a year. In only 24% of the cases, a formal SC was not operational. In two cases, the SC has been suspended, due to political reasons (PARADDER in Mali) or in response to unsatisfactory project implementation (PRODEL in Mozambique). The Health/HIV project to support CNCS in Mozambique does not have a Steering Committee. GIZ has informal bi-monthly technical meetings with CNCS and they hold special meetings to discuss the yearly operational plan. In Palestine, seven of the 12 projects/programmes with a DC agreement do not have a formal SC, but in most cases, another high level committee has been established to discuss progress and challenges. KfW and AFD did not have SCs in Palestine, but this was related to the fact that they delegated implementation to a government agency. An SC was thus regarded as unnecessary. One of these programmes in Gaza, where the Palestinian Authority has limited influence, is not steered by an SC but has, for political reasons, an advisory committee, implementing similar functions. The GBS operations (supported by a TA) in Ghana and Mozambique, both have a formal SC for the coordination and dialogue between the Government and the GBS donors. In Ghana, the SC was led by the Ministry of Finance & Economic Planning on behalf of the Government. Annual reviews have taken place based on a mutually accepted Performance Assessment Framework, agreed with and owned by the Government. In the case of the GBS in Mozambique a Joint SC was established composed of representatives of the Government and of the Programme Aid Partners. A troika participates in the SC on behalf of the Partners. The Troika meets regularly with the Government in the SC, in order to jointly discuss and agree on key strategic issues related to the implementation of the MoU and the monitoring of the Performance Assessment Framework. This analysis of active involvement of the government in SCs also points to a difference per country: for Nicaragua, Tanzania and Mozambique, the documentation and interviews suggests a relatively strong involvement of government actors in the implementation of activities. In one project in Mozambique, however, the active involvement of the lower level government was disappointing, mainly due to capacity limitations; for Timor-Leste, the overall country context is characterised by weak government capacity, politicisation resulting in reorganisations and frequent changes of staff and lack of country-led donor coordination. Project aid with limited country ownership is the dominant aid modality, and the DC projects are no real exception. Implementation reports highlight the lack of country ownership as an explanation of the delays in the execution of the programme. In Benin, the structurally weak capacity of the central government has also been assessed as a limitation for government ownership. There seems to be no specific difficulties with fragile states: in Mali and Palestine, the authorities appear to be as much involved as in other countries, although in the case of Palestine formal SCs have only been established in seven of the 12 cases, but this was also because of the working ways of AFD and KfW – the government agency is largely made responsible and therefore a SC is not relevant. For the twelve TAs related to the GCCA, the available documentation suggests that the partner governments were involved in the design of the projects concerned. Findings from the questionnaire In the questionnaire, a convincing majority of 89% of the respondents indicated that partner governments have been involved in the implementation of DC projects. This involvement might mean a role of the Government either in the project implementation itself or in the SC.

43


Figure 3.2.3 EUD questionnaire response: The partner country was involved in the implementation of DC project(s) / programme(s)


0% 20% 40% 60% 80%

%

Strongly agree 9 26

Agree 22 63

Disagree 4 11


No opinion 0 0


The involvement of the Government in monitoring and evaluation is less convincing with 72%, but here the large share of 17% of the respondents with ‘no opinion’ is striking. Figure 3.2.4 EUD questionnaire response: The partner country took/takes part in the monitoring and evaluation of the DC project(s) / programmes


0% 20% 40% 60% 80%

%

Strongly agree 9 26

Agree 16 46

Disagree 4 11


No opinion 6 17


3.2.6 Reduction of number of active donors (JC 2.4)

JC 2.4. DC has reduced the number of donors directly active in the sector, which has contributed to strengthened ownership and leadership of the partner country as regards the DC project/programme and the related sector policy formulation and implementation

This JC assumes a positive relationship between fewer donors and stronger partner country government ownership. The analysis focusses on whether: (i) The DC agreement has reduced the number of donors active in the sector; (ii) The reduction of the number of donors per sector has strengthened ownership and

leadership of the partner country government. The results of the analysis of the reduction of number of donors active in the sector have already been presented in the section dealing with JC 1.5 and will not be repeated here. JC 1.5 concluded that the DC agreements did not lead to a reduction of the number of donors per sector, because in most cases the DC partners remained active and did not

44


become silent partners. In a limited number of projects and programmes (11 out of 46) a strong (in three projects) or modest (eight projects) positive effect has been observed. Findings from the literature review The literature on aid effectiveness observes that a limited number of donors can have a positive but also a negative effect on partner country’s ownership. Some studies observe a positive link between a reduced number of donors and accountability, which could be described as ‘taking responsibility for development of the sector’. Knack and Rahman (2007) suggest that the feeling of accountability of donors will increase when their number is reduced and a single or dominant donor will have a strong incentive to accept responsibility for success or failure of aid delivery in the sector concerned52. With many donors present in one sector, responsibilities become blurred and no single donor can be held responsible for particular outcomes. Donors will then have an incentive to opt for funding activities serving donor-specific goals rather than the overall development of the sector. Dreher and Michaelowa (2010) state that when a few donors are given the responsibility to lead the support provided to a certain sector on behalf of the donor community, the donor community is accountable as a whole rather than the individual donors – which reduces accountability of separate entities, but strengthens the accountability of the lead donor(s)53. Moreover, a limited number of donors provide a substantial amount of aid to the sector – and when they operate also on behalf of other passive donors e.g. on the basis of DC – they are expected to feel a co-responsibility for the development of the entire sector. If this sense of responsibility is combined with attaching high importance to ownership and leadership of the partner country government, these donors may strengthen that ownership and leadership. With a limited number of donors, it will also be easier for the partner government to coordinate the aid and to lead the sector development process. A potential risk of having only a few active (large) donors on partner country ownership is that donors may become very influential and dominant - due to the volume of their financial support and human resources made available to support policy and strategy formulation and monitor the implementation. Moreover, with only a few donors, there is a higher risk of donors teaming up. Hence, this dominance of a few donors may have a negative effect on the ownership and leadership of the partner country government. These pros and cons of a limited number of donors illustrate the delicacy of identifying an ‘optimal’ number of donors in a sector. This optimal number will very much depend on the country and the sector concerned as well as the mode of operation of the donors and the attention they pay to coordination and promoting country ownership and leadership. Findings from the case studies In JC 1.5, the positive relationship between DC and fewer donors has been rejected, supported by evidence from the case studies. Hence, it is not useful to assess the second part of this JC whether fewer donors have contributed to strengthened leadership and ownership. Findings from the questionnaire The questionnaire confirms the findings from the case studies that DC has not contributed to a reduction in donors, which has strengthened the ownership and leadership of the partner government. The statement is rejected by 65% of the respondents (23 out of 35)

52

Knack, S., and A. Rahman (2007), Donor Fragmentation and Bureaucratic Quality in Aid Recipients. Journal of

Development Economics 83: 176-197. Cited in Iñaki Aldasoro, Peter Nunnenkamp, Rainer Thiele, Less aid proliferation

and more donor coordination? The wide gap between words and deeds, Kiel working paper no 1516, April 2009. 53

Axel Dreher and Katharina Michaelowa, Methodology to measure progress towards in-country division of labor, GTZ

study, November 14, 2010.

45


who disagree or strongly disagree, while only 18% of the respondents (six out of 35) agree or strongly agree with the statement. Figure 3.2.5 EUD questionnaire response: DC has caused a reduction of donors active in the sector, which has strengthened ownership and leadership of the partner country government

Answer Total % response

0% 20% 40% 60% 80%

%

Strongly agree 3 9

Agree 3 9

Disagree 18 51


No opinion 6 17


46


3.3 EQ 3: Complementarity and Added Value

EQ 3: To what extent has Delegated Cooperation strengthened complementarity and added value of the support provided by the EU and other Delegated Cooperation partners?


Complementarity is defined in the Code of Conduct on Complementarity and Division of Labour as “the optimal division of labour between various actors in order to achieve best use of human and financial resources”. It is a broad concept encompassing several dimensions: “Complementarity starts with coordination, but goes much further; it implies that each actor is focusing its assistance on areas where it can add most value, given what others are doing”54. In this description, the three DC outputs linked to the outcome can be recognised: (i) division of labour (“each actor is focusing its assistance”); (ii) comparative advantage (“where it can add most value, given what others are doing”) and (iii) donor coordination and harmonisation (“complementarity starts with coordination”). See reconstructed IED, figure 1.2 in Volume 1. The two outcome level factors ‘strengthening complementarity and increasing added value’ have been put together in one EQ because both factors are intertwined and their evaluation has to be based on more or less the same judgement criteria. Added value as defined and described in the DC documents, refers to the benefits the EU or another donor could contribute to a project or programme in order to enhance achievement of the objectives of the project or programme. This definition is different from EuropeAid’s evaluation criterion “community added value”, which is defined as the additional benefits the EU support will generate compared to “what would have resulted from MS interventions only in the partner country”. That definition is too restrictive55 and does not correspond with how the term ‘added value’ is used in the DC documents.


Surprisingly, strengthened complementarity and added value is not listed as one of the five strategic objectives of Delegated Cooperation in the revised DC Guidelines of 2012. Nevertheless, this outcome cannot be omitted as it was one of the main considerations for launching the DC funding modality in 2007/2008. In the Code of Conduct on Complementarity and Division of Labour (2007), DC was proposed as one of the instruments (called “guiding principles”) to implement that Code. In addition, strengthening complementarity and increasing the added value of the EU and other DC partners (mostly EU Member States or agencies from the Member States) fit into the broader political EU agenda of harmonising development cooperation strategies of the EU and its Member States and strengthening the (policy) coherence within the EU. The assessment of the effects of DC on strengthening complementarity and increasing added value (envisaged DC outcomes) is based on measuring the results of the three

54

Communication from the Commission to the Council and the European Parliament – COM (2007) 72 final, February

2007. 55

As it focuses specifically on the added value of the EU compared to what MS would have contributed, while the term

“added value” in the context of DC refers to the total contribution of the EU and/or other partners to the achievement of

the project objectives.

47


related outputs: improved division of labour, better use of comparative advantages and improved donor coordination and harmonisation. The links between these outcomes and outputs are based on the assumption that division of labour is organised according to the comparative advantages of the respective development partners, which will improve the added value of the support of each of them and the complementarity of the entire support provided by the EU and the other DC partners, while improved donor coordination and harmonisation is focussed on improving the division of labour which strengthens the complementarity and increases the added value of the support provided by the EU and other DC partners.


When analysing complementarity it is important to distinguish In the Conclusions of the EU Council as regards the EU Code of Conduct on Complementarity and Division of Labour, a distinction was made between five different dimensions of complementarity, which all interact with each other and must be seen as a whole56:

In-country complementarity, seeking to address aid fragmentation in a sector or a country in order to decrease the administrative burden and transaction costs of both partner countries and donors and to assure that aid is well spread over sectors and regions in the country;

Cross-sector complementarity seeking to address a situation where at country level some sectors receive much more donor support than others, leading to congestion in some sectors and under-funding of other sectors;

Cross-country complementarity seeking to address a situation where at the global level some countries receive in relative terms much more donor support (“aid darlings”) than others (“aid orphans”);

Vertical complementarity seeking to address a situation where global aid initiatives concentrating on one particular sector worldwide, interfere with policies and strategies at national levels;

Cross-modalities and instruments complementarity seeking to make effective use of aid modalities at sector, country and global levels.

It appears that complementarity is described and defined principally in terms of division of labour, which reconfirms the close relationships between those concepts: complementarity is the outcome of good division of labour, which should be based on the comparative advantages of the various donors, and effective coordination and harmonisation. The Code of Conduct itself focuses mainly on the first three dimensions mentioned above, and more particularly on the division of labour between EU donors and on how EU donors can complement each other’s work (in-country, cross-country and cross-sector). The Code formulates the following implementation strategies for those three dimensions57:

1. In-country: increase participation in joint multi-annual programming based on partner countries’ development strategies, and use the EU joint programming framework gradually and voluntary as a pragmatic tool;

2. Cross-sector: review the state of the self-assessments to be made by donors as regards their comparative advantages and to what extent donors focus on sectors where they have a comparative advantage;

3. Cross-country: conduct an EU-dialogue about future engagement and on strategic planning concerning geographic concentration and country priorities.

56

Idem; EU Toolkit for the implementation of complementarity and division of labour in development policy, 2009. 57

EU Code of Conduct on Complementarity and Division of Labour in Development Policy - Conclusions of the Council

and of the Representatives of the Governments of the Member States meeting within the Council, 9558/07, May 2007.

48


When this evaluation refers to complementarity, it is in fact limiting itself to the dimension of in-country complementarity (which includes in fact cross-sector complementarity). Five of the eleven Guiding Principles of the Code of Conduct deal with in-country complementarity; DC is one of them (see the box below). These guiding principles for in-country complementarity underline that the three outputs (division of labour, comparative advantage and coordination and harmonisation) are intertwined and can only be separated/analytical level.


JC 3.1. DC has helped donors in the partner country to implement the division of labour principles, which has strengthened the complementarity and increased the added value of the support provided by the EU and other DC partners.

This JC assesses whether DC has facilitated and promoted the implementation of division of labour principles, which is supposed to have strengthened the complementarity and to have increased the added value of the aid provided by EU and DC partners. The findings and conclusions as regards to what extent DC has improved the division of labour have already been reported under JC 1.1. The highlights of those analyses were:

An analysis of the entire DA portfolio (164 DAs) revealed that the DA instrument has not really been used to implement a division of labour strategy by means of concentrating the provision of aid on certain focal sectors and to phase out support to non-focal sectors or to become a silent partner in such a sector. In fact, 70 DAs (43% of the entire DA portfolio) were funding projects being part of the focal sectors of the EU support. Another 30 DAs (18% of the DA portfolio) were supporting global or regional progress, or consisted of ad-hoc support where division of labour and a focal sector strategy was not really an issue. Only 64 DAs (39% of the total DA portfolio) were funding projects not being part of the focal sectors of EU support. These DAs could have been part of a division of labour strategy to move out of a sector and/or to become a silent donor in those sectors;

Guiding principles for promoting in-country complementarity

Each donor concentrates its support on three focal sectors in a country + general budget support + support to Non-State Actors + plus research and education schemes. Selection of focal sectors to be based on: o comparative advantage of each donor, self-assessed, endorsed by partner

government and recognised by other donors; o identification (by the partner countries) of sectors and areas for increased or reduced

support and of donors remaining engaged in each sector; o identification (by donors and partner countries) of sectors in which each donor will

remain active and from which sector(s) they will withdraw; o assurance of donors to remain engaged in the focal sectors in the long term (minimum

5-7 years).

Exit from non-focal sector and redeploy funds in the three focal sectors and/or General Budget Support, or stay financially engaged in the non-focal sector via a DC arrangement;

Agree on lead donor arrangements for each sector in charge of donor coordination thereby reducing transaction costs;

Make use of a DC arrangement when supporting a non-focal sector considered strategic for the partner country or the donor;

Ensure adequate donor support to sectors of key priority for poverty reduction. At least one active EU donor per sector, maximum 3-5 active EU donors per sector.

Source: Adapted from the EU Toolkit for the implementation of complementarity and division of labour in development

policy (2009), p5, and the Code of conduct on complementarity and division of labour.

49


For the entire TA portfolio (59 TAs), the figures were more or less similar: 21 TAs (36% of the total TA portfolio) were funding projects being part of the focal sectors of the EU’s support. They could have been part of a division of labour strategy of the TA partner, to move out of a sector. Another 18 TAs (31% of the total TA portfolio) were funding projects not being part of the focal sectors of EU support. These TAs could have caused the EU entering a new sector or expanding its support in a non-focal sector (the opposite of what DC was aiming for). Finally 20 TAs (34% of the portfolio) were supporting global or regional programmes, or consisted of ad-hoc support or GBS;

The nine country case studies have shown that within 31 of the analysed 46 DC supported projects, the DC programme(s) had no effect on the inter-sector division of labour, while in seven cases there has been a negative effect. Only in eight cases (17% of 46) a modest or positive effect has been identified;

Among the 35 EUDs having responded to the question to what extent DC resulted in a better inter-sector division of labour, five responded “strongly” and 17 reported “modestly”. This modestly positive perception is not fully in line with the above mentioned assessments made by the Evaluation Team.

The GCCA case study concluded that the contribution of the seven TAs, supporting 12 GCCA projects, to inter-sectoral division of labour was limited. The link between the GCCA-TAs and the EU’s focal sector strategy was not very strong, because that strategy is difficult to apply to TAs supporting a global thematic facility. On the other hand, the GCCA TAs have contributed indirectly to a better intra-sectoral division of labour, because the EU has actively sought to delegate the implementation of certain GCCA projects to implementing agencies in line with the principles of division of labour. In the country cases of Mozambique and Timor-Leste, where the TA-funded projects were executed via a Delegation Agreement, a modestly positive effect was found on intra-sectoral division of labour. The above summarised findings lead to the conclusion that DC has not contributed much to improving the inter-sector division of labour, which was one of the main output objectives of DC. The contribution of a DC induced improved inter-sector division of labour to the outcome of “strengthening complementarity and increasing the added value of the EU and the other DC partners” has thus also been minimal. It is interesting to make the same type of analysis for intra-sector division of labour. Such an analysis shows that the division of labour effects within the supported sectors are more positive. The main findings as regards intra-sector division of labour are:

The nine country case studies have shown that DC had a strong effect on improving the intra-sector division of labour in 16 of the DC supported projects, while there was a modest effect in 18 DC supported projects (together 81% of the 42 DC supported projects having been scored);

Among the 35 EUDs having responded to the question to what extent DC resulted in a better intra-sector division of labour, 13 responded “strongly” and 13 responded “modestly”. This positive perception corresponds with the assessments made by the Evaluation Team.

These findings indicate that promoting intra-sector division of labour has been a clear feature of DC agreements and may have contributed to “strengthening complementarity and increasing the added value of the EU and the other DC partners”. In the questionnaire, the EUDs have also been asked to rank the seven DC outputs in order of the extent to which DAs have contributed to achieving those outputs. The consolidation of the rankings made by the 35 EUDs, which responded to the questionnaire, reveals that ‘division of labour among donors in the respective sector(s) and/or DA supported project/programme’ was ranked fourth. More precisely, four EUDs

50


gave that output the first rank, four EUDs ranked it second, four EUDs ranked it third, six EUDs ranked it fourth, five EUDs ranked it fifth, eight EUDs ranked it sixth and two EUDs ranked it seventh. The same question was asked as regards TAs and it revealed that ‘improved division of labour’ was ranked second. More precisely: five EUDs gave that output the first rank, two EUDs ranked it second, three EUDs ranked it third, two ranked it fourth, two ranked it fifth, three ranked it sixth and none ranked it seventh (18 respondents in total). These responses/rankings indicate that the EUDs did not consider improved division of labour as one of the outstanding outputs of DC, particularly not in the case of DAs. This is more or less in line with the assessments made by the Evaluation Team (see here above). The final conclusions of the above presented data and analysis are:

DC has contributed very little to improving inter-sector division of labour, and the contribution of a DC induced improved inter-sector division of labour to “strengthening complementarity and increasing the added value of the EU and the other DC partners” (outcome objective) was therefore also minimal;

In most cases DC has had a modest to strong effect on improving intra-sector division of labour (thus within the sectors), which has contributed to “strengthening complementarity and increasing the added value of the EU and the other DC partners”.

3.3.5 Increased use of comparative advantages (JC 3.2)

JC 3.2 Assistance provided by the various donors involved in DC was in line with their comparative advantages, which has strengthened the complementarity and increased the added value of the support provided by the EU and other DC partners.

This JC consists of two components, namely: (i) whether the assistance provided by the DC partners was in line with their comparative advantages, and (ii) to what extent making use of the comparative advantages has strengthened the complementarity and increased the added value of the DC support. The first component is thus focussed on assessing one of the envisaged DC outputs, namely increased use of comparative advantages. For the purpose of this evaluation, comparative advantage is terms of a donor having expertise, experience knowledge and a network greater than those of other donors (see the list of definitions for further clarifications). The DC output ‘making increased use of comparative advantages’ is closely related with the output ‘division of labour’ discussed under JC 1.1 and 3.1 (here above), because it would be logic that the division of labour is based on comparative advantages. The second component of this JC 3.2 is focussed on evaluating the link between the output (comparative advantages) and the outcome (strengthened complementarity and added value). The principle that division of labour should be based on comparative advantages, which will strengthen complementarity and increase added value, is also put forward in the correspondence of the Commission to the EU Council on the Code of Conduct on Division of Labour (February 2007), where it is mentioned that “EU donors should make full use of their respective comparative advantages to (i) enhance division of labour, (ii) deepen concentration of activities, and (iii) develop delegated cooperation”58. Comparative advantage is thus not only seen as a means to enhance division of labour, but also to develop delegated cooperation.

58

Communication from the Commission to the Council and the European Parliament – COM (2007) 72 final, February

2007.

51


The above mentioned quotation does not re-appear in the actual Code of Conduct (May 2007), but the concept of making use of comparative advantages is interwoven in most of the Guiding Principles proposed in that Code of Conduct, in particular that donors should focus their aid on sectors where they have comparative advantages. Moreover, although the Code does not define the concept of comparative advantage, it attempts to guide donors in determining their own comparative advantage59. It is emphasized that comparative advantages of a donor should be self-assessed, endorsed by the partner country and recognised by other donors. It is also stressed that comparative advantage is not primarily based on the amount of aid funds, but on a wide range of issues including geographic and thematic expertise. The Code of Conduct mentions a number of criteria which could be used for assessing comparative advantages (see box hereunder) but does not provide specific measurement tools nor absolute standards.

The EU Toolkit published in 2009 repeats the above mentioned criteria for assessing comparative advantage. It adds that specific criteria of particular importance in the partner country could be added. The Toolkit reiterates that there is no common framework for determining comparative advantages. Although the EU approach as regards assessing comparative advantages is thus quite subjective, a couple of studies have attempted to analyse comparative advantages using objective data. Kharas has calculated the “Revealed Comparative Advantage (RCA)” of each donor, which is an indicator measuring the concentration of aid of a donor aid across countries and sectors60. A high RCA reflects a high level of concentration of aid of the donor concerned in one country or sector, which is interpreted as reflecting a comparative advantage of that donor as regards that country or sector. The study discerns a greater tendency of donors to concentrate their aid on selected countries than on specific sectors. Concentration on selected countries is usually based on historical or geopolitical considerations. On the other hand, most donors have a low RCA as regards sectors; most donors are active in many sectors. This makes in-country coordination complex, while donor agencies do not develop specific expertise in a sector and remain generalists61. Another study carried out by Kharas and Birdsall confirms the results of the 2009 study of

59

EU Code of Conduct on Complementarity and Division of Labour in Development Policy - Conclusions of the Council

and of the Representatives of the Governments of the Member States meeting within the Council, 9558/07, May 2007. 60

Homi Kharas, Development Assistance in the 21st Century, Contribution to the VIII Salamanca Forum, the Fight

Against Hunger and Poverty, July 2-4, 2009. 61

Ibidem.

A donor’s comparative advantage can be determined by, inter alia, any of the following criteria:

presence in the field;

experience in the country, sector or context;

trust and confidence of partner governments and other donors;

technical expertise and specialization of the donor;

volume of aid, at country or sector level;

capacity to enter into new or forward looking policies or sectors;

capacity to react quickly and/or long term predictability;

efficiency of working methodologies, procedures, and quality of human

resources;

relatively better performance - without necessarily absolute advantage;

lower cost compared to other donors with adequate standards of quality;

building new experience and capacities as an emerging donor.

Source: EU Code of Conduct on Complementarity and Division of Labour

52


Kharas. They concluded that the cross-sector RCA of a number of large donors with a mandate to work in various sectors, including the United Kingdom, United States, the International Development Association and the EU, was below the average of this RCA indicator of a series of donors62. This indicates that large donors are not inclined to focus their aid on sectors in which they have a comparative advantage and/or these large donors do not have comparative advantages at sector level. In other studies, the degree of aid fragmentation – which is the mirror image of the concentration measured by the RCA - is seen as an indicator whether or not donors base their support on comparative advantage. It is reasoned that optimal use of comparative advantages means that donors specialise in a limited number of sectors, and thus aid fragmentation should be relatively low and/or decreasing. In a publication of Ohno (2013), it is observed that “the situation where all donors do the same thing regardless of what they can do best, should be avoided. But that is exactly the trend that has been observed in the recent past”63. Various studies conclude also that the decision to support a country is not always entirely based on a ‘rational’ assessment, but is also influenced by strategic and political considerations. This may also be true for the selection of sectors within a country. Making (better) use of comparative advantages is one of the (operational) objectives of DC. In the DC Guidelines of 2007 it is said that the donor managing DC funds should have “well established and recognised technical and financial management capacity to lead on behalf of other donors and is willing to take on this role“, which could be considered as one of the components of comparative advantages64. This statement (condition) is reiterated in the DC guidelines of 201265. The report on Delegated Cooperation (2007-2012) concludes that “the decision on DC should be justified by tangible comparative advantages, aid efficiency and delivery capacity”66. However, none of these documents provide specific guidance on how to determine comparative advantages. Findings from the case studies In the majority of the Action Fiches of the 44 DAs and 40 TAs being part of the case studies of this evaluation, a brief analysis of the experience and expertise of the DC partner is presented, but formal assessments of comparative advantages have not been made. The Action Fiches and other documentation suggest that in some cases a detailed assessment of comparative advantages was also not really needed, because there was only one obvious partner, e.g. a donor or agency having clearly the lead in providing aid to a sector and/or having historical ties with a sector or country (e.g. Portugal in Timor-Leste and AECID in Nicaragua). In the last mentioned example, the partner country government had a strong say in selecting AECID in view of its expertise and experience in Nicaragua. The case studies revealed also that comparative advantage considerations were usually focussed on the project level, rather than on the sector level or on division of labour across sectors. The DA and TA Assessment Forms, which had to be filled in for each DA and TA, could have provided more information on comparative advantage, because they contain a question on comparative advantages and another one on the technical and financial management capacity of the DA partner. However, before 2012 only few of those fiches

62

Nancy Birdsall and Homi Kharas, Quality of Official Development Aid Assessment, Centre for Global Development

(2010). 63

Kenichi Ohno, Izumi Ohno (editors), Eastern and Western Ideas for African Growth: Diversity and Complementarity in

Development Aid (Routledge, 2013). 64

Guidance on joint co-financing with Member States and other bilateral donors (2007), p.4. 65

Guidance paper on Delegated Cooperation (2012), p.4. 66

Report on Delegated Cooperation 2007-2012, p.14.

53


have been filled in (after 2012 more use was made of those fiches). For the 46 DC supported projects, only 13 DC Assessment Fiches have been found. The findings of the nine country case studies as regards the question whether DC has promoted the increased use of the comparative advantages and specific expertise of the EU and the DC partners were quite positive. The case studies concluded that the effects were strong in the case of 29 of the analysed 46 DC supported projects (63%) and modest in 16 cases (25%). Only in one case a negative effect has been reported. See table 3.3.1 for the details. Table 3.3.1 Case study scores: Effects of DC on increasing the use of comparative advantages and specific expertise


Modest effect

No change Negative

effect Total

Benin 5 2 0 0 7

Ghana 3 0 0 0 3

Mali 3 1 0 0 4


Nicaragua 3 1 0 0 4


Tanzania 2 1 0 0 3


Haiti 3 0 0 0 3




The GCCA case study concluded that the contribution of the seven TAs supporting 12 GCCA projects, to increasing the use of comparative advantages was limited. The limited (financial) support from the EU Member States illustrated that the EU was not recognised as the leader in climate change having a clear comparative advantage as regards that issue. The five MS that had signed GCCA-TAs did so because of specific policy considerations. For example Sweden wanted to show its commitment to working on environmental issues in a broader EU context and its adherence to the principles of aid effectiveness. Cyprus and Estonia wanted to make use of the comparative advantage of the EU being a big institution able to manage large projects. In none of the cases, the comparative advantage of the EU in terms of climate change expertise was a fundamental reason for concluding a TA. Besides that, as the EU was aware of its limitations, it delegated the implementation of six GCCA projects to another party having a comparative advantage as regards the project concerned in a particular country, which can be considered as a positive contribution towards making better use of comparative advantages. Findings from the questionnaire Also the EUDs were quite positive about the link between DC and increased use of comparative advantages: 15 (43%) of the 35 EUDs that responded to the questionnaire had the opinion that DC has had a strong effect on increasing the use of the comparative advantages of the EU and the DC partners, while 14 (40%) reported a modest effect. See figure 3.3.1 hereunder.

54


Figure 3.3.1 EUD questionnaire response: To what extent has Delegated Cooperation led to increased use of the comparative advantages of individual donors?


0% 20% 40% 60% 80%

%

Strongly 15 43

Modestly 14 40

No change 5 14

Deterioration 0 0

Don't know 1 3


In the questionnaire, the EUDs have also been asked to rank the seven DC outputs in order of the extent to which DAs have contributed to achieving those outputs. The consolidation of the rankings made by the 35 EUDs having responded to the questionnaire, reveals that ‘increased use of comparative advantages was ranked first. More precisely, eight EUDs gave that output the first rank, seven EUDs ranked it second, five EUDs ranked it third, six EUDs ranked it fourth, four EUDs ranked it fifth, three EUDs ranked it sixth and none ranked it seventh. The same question was asked as regards TAs and it revealed that ‘increased use of comparative advantages was ranked second. More precisely: one EUD gave that output the first rank, eight EUDs ranked it second, two EUDs ranked it third, one EUD ranked it fourth, three EUDs ranked it fifth, two EUDs ranked it sixth and none ranked it seventh (18 respondents in total). Thus the EUDs considered increased use of comparative advantages as being one of the key outputs of DC, particularly in the case of DAs. This confirms the figures reported in table 3.3.1 here above and is line with the conclusions of the case studies. Contribution to strengthening complementarity and increasing the added value DC has clearly led to increased use of comparative advantages of the DC partners, which could make an important contribution to strengthening complementarity and increasing the added value of the EU and the DC partners (the latter will be analysed at EQ level in the main report). The increased use of comparative advantages is closely linked with the improved intra-sector division labour, which means that the increased use of comparative advantages has materialised mainly within the sectors supported by the DC agreements.

3.3.6 Improved donor coordination and harmonisation (JC 3.3)

JC 3.3. DC has led to improved donor coordination and harmonisation, which has contributed to strengthened complementarity and increased added value of the EU and other DC partners, in particular as regards the DC project or programme and/or sector concerned.

This JC assesses the role of DC in improving donor coordination and harmonisation, and how this has contributed to strengthening complementarity and increasing the added value of the EU and other DC partners. “Coordination and harmonisation” are broad and often used terms in the debate about aid effectiveness. Harmonisation was one of the five main principles of the Paris Declaration (2005), to be monitored and evaluated on the basis of the following indicators: (i) use of

55


common arrangements or procedures: percentage of aid provided on the basis of programme-based approaches, and (ii) shared analyses: joint field missions and/or country analytic work. In the Accra Agenda for Change (2008) the term “harmonisation” was less prominent than in the Paris Declaration. The Accra Agenda focused more on coordination: “Donors should improve their co-ordination and avoid duplicated efforts – thereby lowering costs”. The Capacity Development for Development Effectiveness facility explained that the idea of harmonisation from the Paris Declaration was de-emphasized in the Accra Agenda, as it turned out that moving to fully joint initiatives with a lot of donors resulted in additional administration and transaction costs, which diminished the effect on aid efficiency. Therefore, the emphasis on harmonisation was gradually replaced by a focus on a better division of labour67. The Busan Agreement did also not refer to “harmonisation”, but rather to complementarity of actions to reach common goals. The latter is also reflected in the EU Agenda for Change (2011), which advocates single joint programming documents per partner country – which is a coordination instrument – rather than putting a lot of emphasis on harmonisation. The DC Guidance papers contain little direct reference to donor coordination and harmonisation, but many of the individual guidelines deal inherently with coordination and harmonisation, such as “DC should be considered in order to (i) deliver aid more efficiently by sharing and maximising use of technical and financial management capacity and systems, and (ii) encourage use of common monitoring, evaluation and accounting procedures in order to reduce the number of separate donor reviews and accounting procedures”68. These issues are also mentioned in the Guidance paper of 201269. Moreover, DC is supposed to go hand in hand with (joint) co-financing which is intrinsically related with coordination and harmonisation, and DC should be based on division of labour, which requires close coordination. This JC requires that first the effectiveness of the donor coordination and harmonisation mechanisms in the sector of the DC project/programme and of the DC project/programme itself is assessed. Donor coordination refers to donors taking into account each other’s strategies, programmes and projects when defining their own strategies, programmes and projects, while donor harmonisation looks at cooperation between donors aimed at using the same strategies, approaches and procedures for monitoring, evaluation, accounting and procurement when providing development aid. The second step of the analysis is to assess how the improved coordination and harmonisation has contributed to strengthened complementarity and increased added value of the aid provided by the EU and the DC partners. Although the JC considers donor coordination and harmonisation of all donors in the relevant sector, it will particularly reflect on the degree of coordination and harmonisation of the EU and its Member States. Findings from the case studies The findings of the nine country case studies as regards the question whether DC has improved donor coordination and harmonisation were modestly positive. The case studies concluded that the effects were strong in the case of only five (11%) of the analysed 46 DC supported projects and modest in 26 cases (57%), while ‘no change’ was recorded for 15 cases (33%). See table 3.3.2 for the details.

67

Aid Effectiveness Portal, Themes, Paris Principles, “Harmonisation”, http://www.aideffectiveness.org/The-Paris-

Principles-Harmonisation.html, accessed 4 September 4 2015. 68

Guidance on joint co-financing with Member States and other bilateral donors (2007), p.4. 69

Guidance paper on Delegated Cooperation (2012), p.4.

http://www.aideffectiveness.org/The-Paris-Principles-Harmonization.html

http://www.aideffectiveness.org/The-Paris-Principles-Harmonization.html

56


Table 3.3.2 Case study scores: Effects of DC on improving donor coordination and harmonisation


Modest effect

No change Negative

effect Total

Benin 0 5 2 0 7

Ghana 0 2 1 0 3

Mali 1 2 1 0 4


Nicaragua 0 0 4 0 4


Tanzania 0 3 0 0 3


Haiti 0 3 0 0 3




The GCCA case study concluded that the 12 GCCA projects funded by seven TAs had a positive effect on donor coordination and harmonisation at project and sector level. This conclusion was supported by the findings of the global GCCA evaluation. The contribution of the TAs to better coordination at the overall level of the GCCA facility was less clear. Only five MS contributed to the GCCA and the total amount was not very large, which has not facilitated setting up an effective coordination structure and/or promoting effective EU donor coordination and harmonisation in the field of climate change. The latter is not surprising, given the overall difficulty of coordinating aid in the climate change sector. Findings from the questionnaire The EUDs responding to the questionnaire were more positive about the link between DC and improving donor coordination and harmonisation than the case study conclusions: 11 (31%) of the 35 EUDs had the opinion that DC has had a strong effect on improving donor coordination and harmonisation, while 18 (51%) reported a modest effect. See figure 3.3.2 hereunder. Figure 3.3.2 EUD questionnaire response: To what extent has Delegated Cooperation resulted in improved donor coordination and harmonisation in the respective sector(s)?


0% 20% 40% 60% 80%

%

Strongly 11 31

Modestly 18 51

No change 6 17

Deterioration 0 0

Don't know 0 0


In the questionnaire, the EUDs have also been asked to rank the seven DC outputs in order of the extent to which the DAs have contributed to achieving those outputs. The consolidation of the rankings made by the 35 EUDs, which responded to the questionnaire, reveals that ‘improved donor coordination and harmonisation’ was ranked

57


third. More precisely, seven EUDs gave that output the first rank, two EUDs ranked it second, eight EUDs ranked it third, six ranked it fourth, five ranked it fifth, two ranked it sixth and one ranked it seventh. The same question was asked as regards TAs and it revealed that the overall ranking of ‘improved donor coordination was fourth. More precisely: two EUDs gave that output the first rank, three EUDs ranked it second, four EUDs ranked it third, five EUDs ranked it fourth, one EUD ranked it fifth, one EUD ranked it sixth and one EUD ranked it seventh (18 respondents in total). Thus the EUDs considered that DC had a moderate effect on ‘improving coordination and harmonisation’, which more or less confirms the figures presented in figure 3.3.2. Contribution to strengthening complementarity and increasing the added value. DC has thus contributed only modestly to improving donor coordination and harmonisation within the DC-supported sectors and/or programmes, which implies that the contribution of improved donor coordination and harmonisation to ‘strengthening complementarity and increasing the added value’ has also been modest. The scope, depth and intensity of donor coordination and harmonisation in the DC-supported sectors depended mainly on what already existed or what was being promoted through actors and factors other than DC. DC is not an instrument pre-eminently designed for improving donor coordination and harmonisation, but it can modestly advance existing coordination and harmonisation mechanisms, in particular as regards the working relationship between the two DC partners, and in case the DC agreement is based on a joint co-financing arrangement.

58


3.4 EQ 4: Aid Fragmentation

EQ 4: To what extent has Delegated Cooperation reduced the fragmentation of aid?


This EQ is focused on assessing to what extent DC has reduced the fragmentation of aid both within the countries where DC projects were implemented and within the sectors of the projects and/or programmes funded by DC arrangements. The answer to the EQ will be based on in-depth research of the DAs and TAs of the selected case studies. In a paper published in 2010, Dreher and Michaelowa defined in-country aid fragmentation as “the dispersion of aid between numerous donors and / or projects within any given developing country”70. This definition can be applied both to aid fragmentation within a country and aid fragmentation within a sector. There are various methods of calculating aid fragmentation in order to establish trends over time and to make cross-sector and cross-country comparisons. For this DC evaluation, the most simple indicator is used which is the amount of aid divided by the number of active donors (a silent partner is a passive donor). Volume of aid is measured in terms of annual commitments or annual disbursements, depending on availability of data71. All publications dealing with measuring aid fragmentation stress that there is no benchmark or target as regards the most efficient and/or effective level of aid fragmentation. It is thus important to analyse trends and to interpret the trends (qualitatively) in view of the specific context in the country and/or the sectors, and the type of donors and their strategies. Authors have questioned definitions of fragmentation based on number of donors and volume of aid, arguing that indices based on the number of projects better reflect actual problems associated with the lack of donor coordination than indices based on aid volumes72 (see also the section dealing with JC 2.4). Another concept, aid proliferation, is closely related with aid fragmentation. Dreher and Michaelowa have defined in-country aid proliferation as “the dispersion of aid from an individual donor over sectors (or projects) in a particular partner country”. Thus, the higher the number of sectors supported by a particular donor, the higher the aid proliferation. It is likely that when several donors in a country have a high level of aid proliferation, the level of aid fragmentation will also be relatively high. When evaluating aid fragmentation, it is therefore useful to consider also the level of aid proliferation.

70


study, November 14, 2010. The definition resented by them was based on a definition presented by Acharya in 2006

and was said to be in line with DAC terminology. 71

More sophisticated indicators are: (i) applying the Hirschmann-Herfindahl index to the provision of aid, which sums up

the squares of the op the shares of each donor in a sector or country, (ii) or Urs Bürcke, who has used on indicator

focussed on counting the number of smallest donors in a sector (or country) providing together 10% of total aid. See:




See for example Knack, Stephen, and Rahman, Aminur (2004), 'Donor fragmentation and bureaucratic quality in aid

recipients'. World Bank Policy Research Working Paper 3186, World Bank, Washington DC.

59


In an OECD paper written by Bürcky and published in 2011, it was concluded that both in-country aid fragmentation and donor proliferation increased considerably from 2005 to 2009. He counted 8.8% more donors per sector on average and 7.3% more sector engagements per donor on average. The increase in sector engagements coincided with a 21% increase in overall Country Programmable Aid73. These findings indicate that the aid allocation behaviour of DAC reporting donors, on average, did not demonstrate any significant trend towards reduced fragmentation or involvement in fewer sectors until 200974. Also the EU did not perform well according to the findings of Bürcky. In 2005 the EU Institutions provided on average aid to 7.7 sectors in the partner countries they were active (covering a sample of 88 countries). That figure has increased to 8.8 in 2009. The average of all researched donors was 5.9 in 2005 and 6.3 in 2009. A recent study from Pietschmann, entitled “Managing diversity”, notes that aiming at a reduced number of donors per sector and/or per country has become less important75. The focus has now shifted to joint programming at the level of partner countries, while there is also recognition of the value added of having a diversity of development actors with different expertise and experience.


This question is asked because reducing the fragmentation of aid was conceived as one of the main planned outcomes of DC. In the 2012 Guidance paper on Delegated Cooperation, it was stated that through DC “it is possible to concentrate the implementation of similar actions in one hand (the fund managing donor). This should result in a reduction of aid fragmentation”. Factors at output level supposed to contribute to the reduction of aid fragmentation were: improved division of labour, more co-financing, larger projects and programmes, and improved donor coordination and harmonisation (see the IED; figure 2.1 in Volume 1). These factors are the subject of each of the four Judgement Criteria of this EQ. In fact, three of the four JCs are already covered in EQ1, i.e. improved division of labour, co-financing and larger projects and programmes. Finally, the improved coordination and harmonisation among donors involved in DC are expected to have a direct impact on reducing overall aid fragmentation76. The JC on improved coordination and harmonisation is covered in EQ2.

3.4.3 Methodological observations and challenges

The methodological observations regarding the assessment of a reduced number of donors per sector have already been presented in EQ1 and a main observation is that there is no ideal number of donors per sector. In addition, the methodological issues regarding co-financing and larger projects and programmes have also been presented in EQ1, while coordination and harmonisation has been dealt with in EQ3. Therefore, the methodological observations related to these outputs/JCs will not be repeated here. The same applies for the presentation of main findings from the literature, case studies and the

73

Country Programmable Aid (CPA) according to the Creditor Reporting System of the OECD. 74




Elena Pietschmann, Managing diversity: what does it mean? An analysis of different approaches and strategies,

German Development Institute, Discussion Paper 19/2014, 2014. 76

As the 2011 OECD Report on aid fragmentation trends points out, reductions in fragmentation can also be achieved

through other methods than Division of Labour and co-financing, such as joint programming. (p.17).

60


questionnaire on these outputs for which we refer to the relevant sections in EQ1 and EQ3. In this EQ we will focus on the specific contribution of the four outputs to the specific outcome of reduced aid fragmentation. Theoretically, this would require establishing a trend over time or comparing the situation before and after the start of the project supported by DC. However, the available information in the project documents and collected in interviews made it problematic to follow this approach. Therefore, the answer has been primarily based on qualitative information (opinions, perceptions, strategic objectives of the partner country and the donor, etc.) and interpretation of the present state of aid fragmentation in the sector (volume of aid, number of donors and scope of the sector).


JC 4.1. Division of labour among donors providing support to the DC project or

programme and/or the sector concerned has been carried out in such a way that

the aid provided to this project or programme and to the sector became less

fragmented.

The results of the initial analyses and the initial findings as regards division of labour and the links between division of labour and DC have already been reported under JC1.1 and to a lesser extent also under JC3.1. This JC4.1 focuses on the question whether DC has contributed to the reduction of aid fragmentation through improving the division of labour. The instruction to use DC in non-focal sectors is based on the definition of aid fragmentation as the amount of aid divided by the number of active donors. Principle 4 of the Code of Conduct on Division of Labour 2007, called ‘Delegated cooperation/partnership’, states that donors can delegate authority to another donor in a sector that is considered strategic for the partner country. It is mentioned that this is additional to the maximum of three sectors in which a given donor is engaged77. The EU toolkit (2009)78 reiterates this thinking and emphasises DC as an alternative to completely exiting a sector: “Alternatively the withdrawing donor may withdraw from an active role in monitoring and dialogue, but may wish to carry on funding for the sector. In that case co-financing or a silent partnership or delegated cooperation with a donor who remains active in the sector may be a useful solution.” Hence, DC can be used to provide funding to a sector other than the three focal sectors without becoming active in the sector. The number of donors in a sector would decrease, while the total volume of aid provided to the sector was maintained. As a result, the effect of DC on inter-sectoral division of labour is particularly interesting for this JC. Findings As elaborated in JC1.1 and repeated in JC3.1, DC had a limited effect on improving the inter-sectoral division of labour. An analysis of the entire DC portfolio revealed that the DA instrument has not really been used to implement a division of labour strategy by means of concentrating the provision of aid on certain focal sectors and to phase out support to non-focal sectors or to become a silent partner in such a sector. In fact, 70 DAs (43% of the entire DA portfolio) were funding projects being part of the focal sectors of the EU

77

Code of Conduct on Complementarity and Division of Labour (2007), Guiding Principle 5. 78

EU Toolkit for the implementation of complementarity and division of labour in development policy, 2009.

61


support and 18 TAs (33% of the total TA portfolio) were funding projects not being part of the focal sectors of EU support. These results were supported by the case study findings. In 20 of the 31 DA-funded programmes, the EU signed a DA for a programme in its focal sector and in three of the 13 cases received TAs to carry out a programme in an EU non-focal sector. It was also found that when the EU signed a DA in a non-focal sector, it does not necessarily become a passive donor. As a result, the nine country case studies have shown that 31 of the analysed 46 DC agreements had no effect on the inter-sector division of labour, while in seven cases there has been a negative effect. Only in eight cases (17% of 46) a modest or positive effect has been identified. Of these positive cases, two related to DA-funded programmes and six to TA-funded programmes, indicating that TA partners have been more willing to assume a passive role than the EU. Contribution of improved division of labour to reduced aid fragmentation When applying the definition of aid fragmentation as the amount of aid divided by the number of active donors, the limited results on the improvement of inter-sectoral division of labour by DC contributed little to reducing aid fragmentation. This finding has to be placed in the wider context of aid coordination, where broad sector definitions and the various opportunities to provide funding to sectors other than the three focal sectors foster donor proliferation (i.e. donors remain active in multiple sectors). In addition, it was found that for several reasons such as visibility, donors are less interested in a passive role. Nevertheless, the lack of clear instructions on the use of DC, permitting it to be used as well in non-focal sectors, has not helped moving towards a better inter-sectoral division of labour. As mentioned in JC1.1, DC was often used in a pragmatic way to improve division of labour within a sector, which still has a positive effect on transaction costs. However, this issue is less relevant for aid fragmentation as defined in this evaluation, as it does not reduce the number of donors; at best, it prevented overlap within sectors. Therefore, it is important to assess more operational outputs, such as co-financing and larger programmes, in JC4.2 and JC4.3, as this will shed more light on the contributions of joint interventions towards reducing aid fragmentation.

3.4.5 More co-financing (JC 4.2)

JC 4.2. Delegated Cooperation has led to more co-financing of projects and programmes, which has contributed to reduced aid fragmentation

The findings and conclusions of the analysis of the links between DC and co-financing have already been presented in the section dealing with JC1.2. This JC4.2 focuses on the question whether DC has contributed to the reduction of aid fragmentation through more co-financing. The guidance paper on DC (2012) noted that “it is possible to concentrate the implementation of similar actions in one hand (the fund managing donor), thereby avoiding that several EU donors implement such actions in parallel. This should result in a reduction of aid fragmentation and proliferation”79. Co-financing was thus seen as a means to prevent the existence of parallel projects in the same area, i.e. the same volume of aid but a reduced number of fund managing donors.

79

Guidance paper on Delegated Cooperation (2012).

62


Findings JC1.2 describes in detail the findings on more co-financing. It was found that about 75% of the DAs of the entire portfolio were part of a co-financing arrangement (joint and/or parallel); and 54 of the 59 TAs (92%) were co-financed by the EU. These findings were in line with the results from the case studies: 36 (78%) of the DC-supported programmes were co-financed. The number of co-financed arrangement was higher for TA-supported projects (85% of the TA-funded programmes were co-financed against 72% of the DA-funded programmes). Out of the 36 co-financed programmes, 22 were jointly co-financed and 14 were co-financed in parallel. The distinction between joint and parallel co-financing is less relevant here, as this is particularly interesting for transaction costs, but to a lesser extent for reduction of aid fragmentation. Nonetheless, it could be stated that in case of joint co-financing, the relation between the involved donors is stronger while in case of parallel co-financing, donors are often still interested in a certain degree of visibility, which discourages donor concentration / a passive role for the donor. In the 36 co-financed programmes, 19 were financed by the EU and the DC partner. Four of these arrangements consisted of a DA- or TA-funded project with multiple DC partners that provided co-financing. 17 DC-funded programmes were co-financed by the EU, the DC partner and other donors. This indicates that in 21 of the cases, financing from more than two donors was realised, while for 15 programmes, DC only established a programme of the EU and the DC partner. Contribution of more co-financing of projects and programmes to reduced aid fragmentation About three-quarters of the DA-funded programmes were either joint or parallel (or both) co-financed. This number was higher for TAs, which were effectively in all cases co-financed. As such, DC generated more co-financing, which has contributed to reducing aid fragmentation. However, the results for this JC should be seen in the context of the results in the other JCs. Increased co-financing might mean the merging of two programmes which would have otherwise been implemented in parallel, but when this also results in the introduction of a new active donor in the sector, this effect is offset.

3.4.6 Larger programmes (JC 4.3)

JC 4.3. Through Delegated Cooperation the programmes/projects became larger (budget and scope) which has reduced aid fragmentation

For the main findings regarding this output ‘larger projects and programmes’ from literature, the case studies and the questionnaire, we refer to JC 1.3. In this context we will focus on additional qualitative findings that are important to assess the contribution of this output to reduced aid fragmentation. Findings We refer to the scores on larger programmes as presented in Table 3.1.6, showing strong effects in the case of 20 DC projects and programmes (44%) and in another 13 cases modest effects, leading to 75% of the cases with positive effects. Negative effects were limited to four DC projects and programmes, and ‘no change’ occurred in the remaining nine DC projects and programmes. The case studies also pointed at the risk in a minority of cases of too large programmes and projects, which might lead to counter-productive effects on ownership

63


Contribution of larger projects and programmes to reduced aid fragmentation In theory, larger projects and programmes may indeed be a step in the direction of reduced aid fragmentation. EUD respondents to the questionnaire were of the opinion that larger projects and programmes did positively contribute. However, this outcome should be interpreted with care, given the results on inter-sectoral division of labour and reduced number of donors per sector (JC4.1 and 4.2). Reduced aid fragmentation would require a reduced number of donors per sector providing small amounts of support. Larger projects and programmes indeed positively affect the amount of support per donor, but that is only one part of the definition. The EU remains almost always active; for the TA partners, this is true to a lesser extent. For example, the TA partners contributing to the GCCA were predominantly silent partners in the projects. However, in most cases their contribution was limited. Therefore, the evaluation team is of the opinion that larger projects and programmes have a positive, though rather limited, effect on reducing aid fragmentation.

3.4.7 Improved donor coordination and harmonisation (JC 4.4)

JC 4.4. Delegated Cooperation has improved donor coordination and harmonisation, which has contributed to reducing the fragmentation of aid.

For the main findings regarding this output ‘improved donor coordination and harmonisation’ from literature, the case studies and the questionnaire, we refer to JC 3.3. In this context we will focus on additional qualitative findings that are important to assess the contribution of this output to reduced aid fragmentation. Findings We first refer to the scores on improved donor coordination and harmonisation as presented in Table 3.3.2., showing strong effects in the case of only five (11%) of the 46 DC projects and programmes and modest in 26 cases (57%), while ‘no change’ was recorded for 15 cases (33%). This can be considered as an overall modest positive effect. The responses to the questionnaire were even slightly more positive than the scores from the case studies, but in terms of ranking this output was hardly ranked in the first or second place. Thus the EUDs also considered that DC had a moderate effect on ‘improving coordination and harmonisation’. When assessing the contribution of donor coordination and harmonisation to aid fragmentation, the country context is quite important as the levels of aid fragmentation and the degree of coordination and harmonisation differ from one country to another. There are case study countries where traditionally aid has been quite dispersed and where it is very difficult for DC to make a significant contribution to reduced aid fragmentation. Timor-Leste and Palestine are typical countries where aid has been always quite dispersed. In other countries some trends towards less fragmentation were visible for some time, which often went together with a change in aid modalities towards more budget support, basket funds, etc. This has been the case in most case study countries, but it should be noted that in some countries, because of changes in the political situation or an armed conflict, these trends have been reversed. Typical examples to be mentioned here are Nicaragua and Mali. Donor coordination and harmonisation can be assessed at different levels, while the degree of coordination is related to the country context, and in particular to the degree of aid fragmentation and aid governance approaches. The following coordination levels have been distinguished when analysing coordination:

64


1. Donor coordination led by the partner government. In most countries an aid architecture has been established with an overall structure and sector groups that are led by a specific Ministry and co-led by a leading donor in the sector;

2. Donor coordination among donors. Often in addition to the formal aid coordination led by the partner country government, donors have their own coordination and consultation meeting often at sector level where the co-lead donor in the government led coordination has the lead in donor coordination;

3. EU - Member States coordination. The EU Heads of Delegation meet on a regular basis to discuss political coordination and in addition in most countries the EU Heads of Cooperation of the Member States donors present in a country also meet on a regular basis. In addition there might be sector EU-MS meetings, but that is often not the case;

4. Informal coordination among donors and (local) authorities. In addition, the abovementioned three more formal levels of coordination, in most countries there is quite some informal coordination at sub-sector level and in specific regions and/or districts to inform each other and to avoid overlaps.

In practice, respondents in the interviews for the case studies and in their responses to the questionnaire referred to these four levels, but no clear distinction was made in the scoring. At present, the formal aid coordination structures in most case study countries are still present and functioning, but often there appears to be some ‘coordination fatigue’. For example, in Palestine after the Oslo Accords in 1993 a very elaborate aid coordination structure was agreed upon. That structure is still in place, but a recent review indicated that the system has lost functionality, which is related to the “metastasized” peace process80. Of course, this problematic politicised aid coordination context affects to an important extent the contribution DC projects can make to improved donor coordination. Only in a small sub-sector such as Technical and Vocational Education Training, which is supported by only a few donors, and where no formal coordination structure was in place, the DC agreements were able to have a strong effect on donor coordination and harmonisation. As DC involves collaboration between donors and also leads often to bigger and co-funded projects, most DC projects and programmes have a modest effect on donor coordination and harmonisation. In countries with a more or less formal aid coordination context (level 1 according to the classification presented above), other forms of donor coordination become more important such as coordination among donors (level 2), EU-MS coordination (level 3), and informal on the ground coordination (level 4). The configuration per case study country differs, but in most countries some positive effects of DC on donor coordination was found, which explains indeed the modest effects. EU-MS coordination becomes important in politically-sensitive situations where like-minded donors have to operate together in order to achieve political and development goals. It became apparent during interviews at HQ and in the field that the distinction between a donor (funding agency) and an implementing agency should be noticed, as it can at times affect effective donor coordination. Implementing donor agencies are not always engaged in the policy dialogue – in these cases, it depends on the exchange of the implementing agency and its foreign representation in the country how much leverage the donors have in the policy dialogue. The implementing agencies are not deciding on future budget allocations and political decisions; this has to be borne in mind when speaking about how DC improves donor coordination and harmonisation. This difference also affects the DA/TA ratio, as implementing agencies do not decide nor have the funds or authority to engage in a TA. Reciprocity is therefore more difficult (but this will be further explained in EQ7).

80

Nigel Roberts, Reforming aid management in the West Bank and Gaza: Proposals for discussion, 16 December 2015.

65


Contribution of improved donor coordination and harmonisation to reduced aid fragmentation The main question is whether the modest effect on improved donor coordination and harmonisation did contribute to less aid fragmentation. The responses to the questionnaire were mainly positive. However, the interpretation of the contribution of improved donor coordination and harmonisation is not simple and requires clear agreement on concepts and definitions. This has proven to be quite problematic and was also indicated by respondents to the questionnaire. According to the Evaluation Team, the contribution of this specific output to reduced aid fragmentation can be easily overestimated. DC projects and programmes only represent a very small proportion of the total aid volume to specific sectors, and represent also a small proportion of the EU portfolio. Therefore, it is difficult for DC to really affect aid fragmentation at country and sector level. However, that is the level on which the evaluation is focused. Probably in the questionnaire sub-sector levels are also taken into account. The evaluation team assesses that the modest effects recorded regarding improved donor coordination and harmonisation can in most cases not really affect the level of aid fragmentation.

66


3.5 EQ 5: Alignment


Alignment is defined in the Paris Declaration on Aid Effectiveness (2005) as “Donors base their support on partner countries’ national development strategies, institutions and procedures”. This EQ has examined policy and systems alignment in JC 5.1 and JC 5.2 respectively. JC 5.1 has assessed to what extent the DC-funded projects and programmes are embedded in the partner country’s policies and strategies. JC 5.2 has analysed whether the implementation of those projects and programmes is managed and implemented by the country’s institutions and on the basis of their systems and procedures rather than through stand-alone structures. It is assumed that these JCs could be factors contributing to strengthening alignment. In the OECD-DAC Guidelines entitled ‘Harmonising Donor Practices for Effective Aid Delivery’ published in 2003, alignment is not so much considered as an output or outcome of DC but rather as a guiding principle of DC (indicating how to implement DC). Referring to policy alignment, the OECD-DAC put it as follows: “Delegated co-operation arrangements should be consistent with the partner government’s poverty reduction strategy or equivalent national framework, and they should support, not undermine, partner governments’ capacity building and accountability to their own people”81. With respect to systems alignment, the same document recommended that the delegating and fund managing donors should be as flexible as possible in accepting the procedures of the partner country. It is recommended that donors determine the extent to which procedures of DC arrangements can be adjusted to comply with the requirements of the partner government involved82. In order to achieve real benefits from DC arrangements, the collaborating donors should make use of the partner country's administrative systems for accounting, audit, statistics etc. whenever possible. Where it is not possible to use the procedures of the partner country, delegating and fund managing donors should aim for adopting common procedures; the ultimate stage being joint co-financing arrangements. The literature also implicitly suggests that the implementation of these good DC practices – especially with regards to systems alignment - may not depend so much on the DC process per se but rather on the provisions of the fund managing donor as regards allowing to use country systems for implementing projects. In other words, donors having little flexibility to align their project procedures to country systems can do little to strengthen alignment in the context of DC. The Evaluation of the Paris Declaration from 2011 outlined these limitations, both as regards policy alignment and systems alignment83. It observed that overall the alignment of donors’ aid with country priorities had only moderately increased, partly because the strategic priorities of the partner countries were seen by donors as not clear enough. The evaluation report also mentions that programme-based approaches could strengthen policy alignment. As regards systems alignment, this Paris Declaration evaluation concluded that only limited, if any, overall increase in the use of country systems and procedures had been

81

Harmonising Donor Practices for Effective Aid Delivery, DAC Guidelines and References Series, 2003m p.91. 82

Ibid, p.93 and 94. 83

The Evaluation of the Paris Declaration, May 2011.

67


achieved. This was the case with most donors, and notably as regards the use of financial and procurement systems of the partner government, mainly because of lack of confidence by donors in those systems. Furthermore, the reliance on project implementation units was not considered as being a burning issue – partly because it is not easy to make a clear distinction between units operating largely independently from the responsible ministry and units well integrated in such a ministry. More recently, the 2014 Progress Report on Making Development Cooperation More Effective84 found little change in use of country systems since 2010 and a great variation in the use of country results frameworks between providers. Those findings are also confirmed by a recent report, which studied in particular systems alignment (in Africa).85 What emerged from that report - unsurprisingly - is that project support may not be as country-system friendly as other aid modalities (budget support, pooled funding). The report notes that many donors do not have any provision for the use of country systems apart from general or sector budget support, and points particularly at the EU in that respect. In summary, the literature review suggests that the role of DC in strengthening alignment should not be overestimated. If donors do not have provisions for using country systems when implementing projects, there is little DC can do to adjust these provisions for specific DC projects.


Strengthening alignment was and is a key element of the underpinning of the DC policies and strategies. For instance in the EU’s Code of Conduct on Division of labour (2007) it was emphasised that division of labour (which is the main foundation pillar of Delegated Cooperation) should be “embedded in the principles of ownership, alignment, harmonisation, management by results and mutual accountability”. Another reference to alignment can be found in the EU’s Report on Delegated Cooperation 2007-2012, where it was stated that Delegated Cooperation is a “modality for moving towards larger programmes aligned with the partner country’s policy jointly funded with MS and other bilateral donors”. It is therefore relevant to explore and evaluate whether Delegated Cooperation has contributed to strengthening alignment.


The various EU policy and strategy documents as regards DC do not elaborate on how DC should or could contribute to strengthening alignment. Therefore there are no intended causal links which could be subject of evaluation. For the same reason the IED does not show causal links between the outputs and outcome levels. The Evaluation Team has identified two main (potential) outputs which could strengthen alignment, namely policy alignment and systems alignment, which are the subjects of two JCs, 5.1 and 5.2. These potential outputs are derived from the definition/description of alignment presented in the Paris Declaration: ” … the principle of alignment refers to two important changes to aid practice. The first is that donors should base their support on the partner country’s development priorities, policies and strategies (‘policy alignment’). The second is that aid should be delivered as far as possible using country systems for managing development activities, rather than through stand-alone project structures (‘systems alignment’).”

84

OECD, Global Partnership for Effective Development Cooperation, Progress Report on Making Development

Cooperation More Effective, 2014. 85

Towards a Greater Use of Country Systems in Africa, Recent Trends and Approaches, Synthesis Report, August 2014.

68


3.5.4 Embedded in policies and strategies of partner country (JC 5.1)

JC 5.1. Delegated Cooperation projects and programmes were based on and embedded in the policies and strategies of the partner country, which strengthened the alignment of aid.

This JC assesses whether DC projects have contributed to strengthen the alignment of the related projects and programmes to the country’s development priorities, policies and strategies (‘policy alignment’). The analysis of the available documents of the DC agreements of the case studies has shown that most DC-supported projects and programmes have been part of programme based approaches and that many of them have supported one component of a broader sector programme, with other components being financed by other donors. Most of the identification reports and progress reports of the projects and programmes supported by one or more DC agreements refer to their formal alignment with the policies of the Government. This includes a large number of overall, sectoral and thematic strategy and policy documents. Thus, the formal policy alignment of most DC projects is high, but policy alignment only makes sense if the policy or strategy is really owned by the government. The real level of ownership of the partner government for these strategies, however, seems to vary from project to project. The ownership of the partner Government for the policies and strategies has been discussed in EQ2. Findings from the case studies The case studies have paid attention to the policy alignment of the projects and programmes funded by DCs, and specifically to the contribution of the DC agreements to strengthening the policy alignment. A high level of policy alignment of a project or programme does not mean automatically that these DC agreements have contributed to the strengthening of the policy alignment. The findings of the contribution of the DC agreements are shown in the table below. None of the DC projects has contributed strongly to better policy alignment. The assessment for 67% (28 out of 42) of the projects is neutral and modestly positive for the other 33% (14 out of 42), respectively. Table 3.5.1 Case study scores: Effect of DC agreements on strengthening policy alignment


Modest effect

No change Negative

effect Total

Benin 0 3 3 0 6

Ghana 0 0 1 0 1

Mali 0 3 0 0 3


Nicaragua 0 0 4 0 4


Tanzania 0 1 2 0 3


Haiti 0 0 3 0 3


of which DAs 0 11 20 0 31

of which TAs 0 3 8 0 11

* Four projects and programmes (2 in Ghana, 1 in Mali and 1 in Benin) out of 46 have not been scored on alignment

because the government has not or hardly been involved in these projects.

Five of the projects and programmes with a modest effect have been implemented in Palestine. The level of policy alignment in Palestine tends to vary from one sector to

69


another. In the broader sectors, e.g. Water and Sanitation and Private Sector Development, municipal development and local governance, new laws and policies have been formulated, but the necessary reforms are implemented at a very low pace and the policy dialogue is rather limited. For some more narrow sub-sectors, such as TVET, solid waste management and the police, it has been possible to make some progress regarding policy alignment. This might be related to the fact that at sub-sector level, issues are more technical and less politically sensitive. From the perspective of ownership and policy alignment, it is worth mentioning that DCs have not always been very well planned in advance, i.e. not covered the Country Strategy Paper (CSP) and often result from discussions in the country at the operational level. As a consequence, the alignment of these DCs have not been optimal. Examples include the DAs for a project targeting small and medium-sized enterprises (PAFIRIZ) in Benin and for a regional development programme (PARADDER) in Mali. Findings from the questionnaire In the questionnaire, the respondents have given feedback whether the DC projects were based and embedded in the policies and strategies of the partner country. They did not respond whether a DC agreement itself has strengthened the policy alignment (as was assessed in the case studies). With regard to this first question, 91% of the respondents (32 out of 35) supported this statement (strongly agree or agree). This confirms the initial findings from the desk study that the DC projects are generally well aligned and embedded in some form of sector strategy. This also confirms the impression that DCs have been prepared and concluded in a context where a relatively high level of policy alignment was already existing, rather than that the DCs played a crucial role in creating and strengthening alignment. Figure 3.5.1 EUD questionnaire response: The DC project(s) / programme(s) was/were based on and embedded in the policies and strategies of the partner country


0% 20% 40% 60% 80%

%

Strongly agree 14 40

Agree 18 51

Disagree 3 9


No opinion 0 0


3.5.5 Based on national procedures and systems (JC 5.2)

JC 5.2. Implementation of the Delegated Cooperation projects and programmes was managed by the regular implementation entities of the partner country and was based on national procedures and systems, which strengthened the alignment of aid.

This JC assesses whether the aid provided through DC was/is managed by entities of the partner government on the basis of national rules and procedures, rather than through stand-alone project with parallel structures and systems. This is called ‘systems alignment’.

70


As explained in the introduction of this chapter, strong systems alignment means that donors use the national procedures and involve government institutions in project implementation and monitoring. It implies that the government is involved in the implementation, management, coordination and monitoring of the DC projects. Therefore a strong correlation is foreseen between the scores in the case studies with regard to strengthened ownership (EQ 2) and with regard to strengthened systems alignment. Furthermore, there is also a link between this JC and the use of a single management system (JC1.4). In JC1.4 it was explained to what extent one single management system is being used - which can be the DC partner’s system or a partner country system. Findings from the case studies More in-depth investigation in the case studies has shown that the level of effective systems alignment seems to vary, between countries, but especially between DC partners. The level of systems alignment is primarily determined by the differences in their working methods and procedures. In this regard, EQ2 has made a distinction between agencies such as KfW and AFD, which tend to make more use of partner country systems, and agencies such as GIZ, BTC, CICL and AECID, which tend to use their own systems and procedures, parallel to the government ones. The latter work mainly with service contractors and make limited use of national government systems. In projects and programmes managed by these DC partners there is generally less involvement of the counterpart institutions in project implementation than in the case of direct management by the EU. This conclusion is similar with regards the level of ownership and leadership of the partner Government, discussed in EQ2. In 57% (24 out of 42) of the projects and programmes no effect of the DC agreements on strengthening systems alignment has been observed. Two projects (5%) revealed a negative effect and in 16 projects (38%) a modestly positive effect was shown. Table 3.5.2 Case study scores: Effect of DC agreements on strengthening systems alignment


Modest effect

No change Negative

effect Total

Benin 0 2 3 1 6

Ghana 0 0 1 0 1

Mali 0 1 2 0 3


Nicaragua 0 0 4 0 4


Tanzania 0 1 1 1 3


Haiti 0 2 1 0 3




* 4 projects(2 in Ghana, 1 in Mali and 1 in Benin) have not been scored on alignment because the government was not or

hardly involved.

Sixteen projects and programmes with DC agreements have contributed modestly to strengthened systems alignment. Most of these are managed by agencies which use the national systems and procedures. The two programmes in Benin, for instance, have been implemented by the Netherlands Ministry of Foreign Affairs and Danida, and largely used the systems and procedures of the Government of Benin. Most of the seven projects in

71


Palestine have been implemented by agencies such as KfW and AFD, which tend to make more use of partner country systems and involve Palestinian government institutions or agencies as project executing agencies. One other programme is the PEGASE Direct Financial Support to Recurrent Expenditures of the Palestinian Authority. This mechanism is managed by the EU which works closely with Palestinian institutions. Two projects and programmes showed a negative effect on systems alignment. For one project in Benin, the implementing agency (BTC) was using direct management modality with full parallel systems. The counterfactual would have been a stand-alone EU project, using the EU procedures, which would have been more aligned. The negative effect in Tanzania is related an environmental project delegated to BTC. BTC needed to adjust its implementation modality from joint management to direct management (by BTC only), leading to a change in rules and regulations. Initially, the programme was managed according to the government rules, but BTC had to change to direct management in response to the fact that it had not yet been the subject of a seventh pillar assessment. BTC is now managing this project using its own rules and regulations. This project was also discussed in JC 2.4, because this change also had a negative effect on ownership and leadership. Both TAs for General Budget Support in Mozambique and Ghana have been scored ‘no change’. The GBS is fully aligned, but both TAs have not changed this situation. In general, DC arrangements has mostly been used for projects, and only in these two occasions for budget support programmes. However, budget support would provide ample opportunities for DC (in particular through TAs) because of the importance attached to setting up joint budget support operations as set out in the budget support guidelines of the EU and various other donors. In this way, DC operations could contribute to strengthening both policy alignment and systems alignment, because alignment is an inherent characteristic of budget support. With only two GBS operations in the period 2007-2014 benefiting from DC, the attempt of the EU to get Member States to co-finance budget support operations through delegated cooperation has not been very successful. Findings from the questionnaire The feedback from respondents on the question about the use of national procedures confirms the mixed views and picture with regard to the level of systems alignment. From the 35 respondents, 43% (15 respondents) has the impression that DC projects and programmes have largely been based on national procedures and systems, while 55% (19 respondents) disagrees or even disagrees strongly and has a different view. Figure 3.5.2 EUD questionnaire response: Implementation of the DC project(s) / programme(s) was based on national procedures and systems


0% 20% 40% 60% 80%

%

Strongly agree 2 6

Agree 13 37

Disagree 16 46


No opinion 1 3


72


The feedback from respondents on the question about involvement of national implementing entities in the management of projects and programmes shows a similar mixed picture. Among the 35 respondents, 60% (21 respondents) indicate that DC projects and programmes have been managed by the implementation entities of the partner country, while the other 40% (14 respondents) disagree. Figure 3.5.3 EUD questionnaire response: Implementation of the DC project(s) / programme(s) was managed by the regular implementation entities of the partner country


0% 20% 40% 60% 80%

%

Strongly agree 3 9

Agree 18 51

Disagree 13 37


No opinion 0 0


The feedback from respondents on the overall impact of DC agreements on strengthening of alignment of aid was generally positive. About 23% of the 35 respondents (eight out of 35) indicated a strong impact and another 57% (20 out of 35) a modest impact. The other 20% (seven respondents) did not see any impact of DC agreements on strengthening alignment. Figure 3.5.4 EUD questionnaire response: Overall impact of DC on strengthening alignment of aid


0% 20% 40% 60% 80%

%

DC has contributed substantially to strengthening alignment of aid

8 23

DC has contributed modestly to strengthening alignment of aid

20 57

DC has had no effect on the level of alignment of aid

7 20

DC decreased the alignment of aid 0 0

Don't know 0 0


73


3.6 EQ 6: Visibility

EQ 6: To what extent has the visibility of the EU and other participating donors been ensured in case of projects and programmes funded via Delegated Cooperation?


The definition and scope of visibility of the EU has evolved over the years. It moved from a static visibility concept in the sense of logo’s, letter heads etc. towards a dynamic definition of visibility referring to a whole range of communication activities to be undertaken. Visibility is also supposed to contribute to democratic accountability within the EU and mutual accountability towards third country actors. The Thematic Evaluation on the Visibility of EU External Actions 2005-2010 defined visibility as: “The awareness and perception of the image of EU external action among EU and non-EU stakeholders resulting from EU communication activities or from other actions that have an impact on this image”86. The definition of visibility with regard to DC is a bit narrower compared to the definition of the Thematic Evaluation, as the 2012 Guidance on DC refers to “appropriate visibility of each of the involved donors” and the promotion of their role in the DC funded project or programme. The EU has possibly assumed that the visibility of the delegating donor would be weakened and that special measures are required to minimize/mitigate the risk of losing visibility. Those special visibility measures and the results in terms of safeguarding visibility are assessed by this EQ and its JCs. At the level of the JCs a distinction is made between DAs and TAs.


The focus on visibility in this evaluation reflects the increased attention given by donors to the concept in recent years, in a context where political support towards development cooperation is decreasing. Against that background, modalities such as DC could face challenges, unless visibility is properly ensured through concrete actions. The Commission has recently decide to reduce its engagement (through contribution agreements) with United Nations (UN) agencies, partly because of (lack of) visibility concerns. More specifically, an EQ is devoted to assessing the results of visibility activities, because (i) visibility is mentioned specifically in the Terms of Reference and (ii) ensuring visibility became an important element in the DC policy and programming documents from 2012 onwards. Ensuring “appropriate visibility of each of the involved donors” was mentioned as a “general Delegated Cooperation condition” for the first time in the 2012 Report on Delegated Cooperation. In the same report it was concluded that “visibility of the EU in projects and programmes implemented through Delegated Cooperation had generally

86

Consortium PARTICIP-ADE-DIE-DRN-ECDPM-ODI, Thematic Evaluation of the Visibility of EU external action 2005-

2010, June 2012.

74


been satisfactory”. This subject was again brought to the fore in the Guidance paper on Delegated Cooperation of 2012, where it was stated that DC should “result in visibility gains” and “appropriate visibility of the donors involved”. Therefore the issue of ensuring visibility refers to both the EU and the other donors involved. Concerns about visibility will particularly be an issue for the delegating partner; thus a concern of the EU in the case of DAs and a concern of another donor in the case of a TA. In this context it is also worth mentioning that in the Code of Conduct on the Division of Labour (2007), the “promotion of the role of the EU” was mentioned as one of the issues to be considered in the case of DC. This was reiterated in the 2012 Guidance Paper on Delegated Cooperation, with particular reference to TA. Promotion is a wider concept than just visibility, but visibility is surely part of it.


The EU published instructions on ensuring EU visibility in the EU Visibility Guidelines for EU external action issued in 2005. The purpose of those Guidelines was “to ensure that projects that are wholly or partially funded by the European Union (EU) visibly acknowledge the support of the EU”87. In 2010, a new version of the Guidelines was issued entitled “Communication and visibility manual for EU external relations”. That manual broadens the scope of visibility and pays more attention to communication, awareness raising and public engagement than the Guidelines of 2010. The manual is designed to “ensure that actions that are wholly or partially funded by the European Union incorporate information and communication activities designed to raise the awareness of specific or general audiences of the reasons for the action and the EU support for the action in the country or region concerned, as well as the results and the impact of this support”88. The manual does not only set out the requirements and guidelines for briefings, written materials, etc., but also offers tools and instructions for the setup of a communication strategy aimed at highlighting the results of the EU support. Furthermore, it should be noted that visibility is not part of the IED of DC, because it is not one of the envisaged outputs or outcomes contributing to improved aid effectiveness. There are also no linkages between visibility and the other outputs and outcomes of the IED. Ensuring visibility is in fact a kind of operational side objective when implementing a DC funded project or programme. When evaluating visibility, it has also to be checked whether activities aimed at ‘ensuring visibility’ do not undermine achievement of the main outcomes of DC and the overall objective of improving the effectiveness of aid delivery. Moreover, too many visibility related activities could lead to an increase in transaction costs and make aid less efficient. Findings from relevant studies and evaluations In 2012, a Thematic Evaluation of the Visibility of EU external action was published. That evaluation assessed the visibility of the EU external actions between 2005 – 2010 (thus partly before the introduction of the Lisbon Treaty signed in 2007 and the related reorganisation of the EU’s External Action Service). The main findings of the evaluations are summarised in the text box below.

87

EC, EU Visibility Guidelines for EU external action, September 2005. 88

EC, Communication and visibility manual for EU external actions, 2010.

75


Summary of the main findings of the Thematic Evaluation of the Visibility of EU external action 2005-2010

The fourth conclusion of that evaluation refers to “working in partnership with others” and may thus also be relevant to DC. More precisely, that evaluation concluded that: “Partnerships with other organisations (UN, governments, NGOs, etc.) are vital in EU external action, but there is a trade-off as visibility needs then to be shared. This remains a source of tension and too much insistence being put on EU visibility can undermine the effectiveness of the cooperation and the sense of ownership felt by partners.” A paper written by Holger Mürle (German Development Institute, 2007) has made a causal connection between division of labour and visibility. The paper conceives visibility as creating a single European identity that should be spread by the EU and all its Member States. it states that the “progress towards a better internal division of labour would increase the visibility and the political influence of the EU in international development co-operation”89. Delegated Cooperation could contribute to a “unified” EU image through improved cooperation and division of labour – although the paper refers to DC as a “less advanced form of Division of Labour”.

3.6.4 Visibility of the EU in Delegation Agreements (JC 6.1)

JC 6.1. Concrete actions have been taken to ensure the visibility of the EU in the case of Delegation Agreements.

This first Judgement Criteria (6.1) of EQ6 deals with DAs while the next one (6.2) cover the TAs. They are discussed separately because there might be specific issue related to both of them. The template for DA Assessment Fiches90 includes two visibility questions, namely”:

“What will be the impact of the delegated action as regards EU visibility?”

“What are the overall commitment and specific actions envisaged to guarantee the visibility of the EU?”

89

Holger Mürle, Towards a Division of Labour in European Development Cooperation: Operational Options, German

Development Institute, Discussion Paper 6/2007. 90

Both the version of 2012 and the version used prior to 2012.

The evaluation concluded that:

1) The image of the EU external action has generally been in-line with pre-

Lisbon official priorities although the quality of EU's image could be improved

notably by communicating more on results;

2) The actual conduct of communication activities based on clear political

strategies by the EU is reasonably good, though far from perfect and lacking

clear leadership and coordination, positive signs of progress are expected

notably with the definition of clear political priorities;

3) Considering the EU's heterogeneous image, the EU needs to manage

sensitively the geographic and constituency variations;

4) Working in partnership with others is essential but involves a trade-off in

terms of EU visibility and a coordination between partners; and

5) The resources to promote the visibility of EU external action have been

adequate.

Source: Consortium PARTICIP-ADE-DIE-DRN-ECDPM-ODI, Thematic Evaluation of the Visibility of EU

external action 2005-2010, June 2012.

76


The template attached to the Companion 4.1 (September 2014; Indirect Management Delegation Agreement (IMDA) template) has simplified these questions to: “describe how EU visibility will be ensured.” The wording of these questions makes clear that in case of DAs that the emphasis is put on (i) ensuring visibility rather than on increasing visibility, and on (ii) EU’s visibility while the visibility of the DA partner is not mentioned, presumably because the EU is concerned that it will be insufficiently visible in case of a DA. Article 6 of the General Conditions to DAs is devoted to promoting EU visibility. That article has 10 sub-articles which are all related to ensuring that project equipment and project outputs can be identified as having been funded/supported by the EU, such as displaying the logo and publicising the amount of the EU contribution. Furthermore, one sub-article stipulates that: “The Delegatee body shall conduct regular checks at least once a year in order to ensure whether the requirements related to visibility of the Action are duly respected.” Annex 1 of DAs contains the description of the Action. In that description, a chapter should be devoted to the approach towards communication and visibility. Furthermore, a comprehensive communication and visibility plan has to be drafted during the inception phase of the project. Findings from case studies In the following table the scores regarding 44 DAs related to 33 projects and programmes are summarized, which shows that in nearly all project and programmes (31/33), some actions were taken to ensure visibility of the EU. There were only two projects were no action was taken to ensure visibility. The main finding is therefore that concrete action has consistently been taken to ensure the visibility of the EU in DA projects, albeit to different levels according to countries and DA related projects. Table 3.6.1 Case study scores: Extent to which actions have been taken to ensure the visibility of the EU and the DA partners

Country Yes Partly No Total

Benin 2 3 0 5

Ghana 1 1 0 2

Mali 3 0 0 3

Mozambique 1 2 0 3

Nicaragua 2 1 1 4

Palestine 3 5 1 9

Tanzania 2 0 0 2

Timor-Leste 3 0 0 3

Haiti 0 2 0 2


This broadly positive finding reflects the fact that DA partners have generally complied with the key visibility requirements as stipulated in DAs’ General Conditions, notwithstanding a few exceptions. Evidence from case studies suggests that the EU has generally been attentive to those requirements – the EU did not want to lose visibility as a result of DC – and that DA partners, even if there are some cases where they found those slightly excessive, have applied those in a consistent manner. For nearly all DA related projects, logos of the EU are shown on reports and communication material and often there is a specific text mentioning who is funding the project or activity. The EU is also invited to attend the main related events. The evidence on the preparation of Communication and Visibility plans, as also required by the DA’s General Conditions, is

77


less conclusive. While it could be determined that in many cases, such plans have been drafted, there were also other cases where this had not been done or DA partners and the EU were not aware of the existence of such plans. Some DC partners viewed the requirement of a DA communication plan as excessive in a context where the management of the DA was already relatively time-consuming. In the case of the TECNICA project in Nicaragua, the EUD was not satisfied with the level of visibility attributed to the EU by the AECID, whereas the latter claims that this was due to the fact that there were delays in the approval of the Communication and Visibility Plan. In some cases, (lack of) visibility was seen as an issue in the early phases of the DA project, but this was subsequently resolved through dialogue between the DA partners. In Mozambique, visibility has been an important point of discussion between the EU, Danida and Ireland with regards the Environment Sector Programme Support (ESPS II) project, as Ireland and the EU were hardly visible, however consensus has been reached on how to ensure more visibility of both partners. In certain DA projects, there was a tension between visibility requirements and aid effectiveness principles. While this was not observed across DA projects, there were some cases where visibility reasons have pushed the EU to fund specific projects components (i.e. in Palestine). In other cases, tensions arose because the implementing agencies sub-delegated implementation tasks to country partner authorities. This contributed to systems alignment, but the EU felt that this went at the expense of its visibility. Meanwhile, visibility requirements have led to increased transaction costs for DA partners, but those appear to have been (in most cases) relatively manageable. Visibility requirements have generally been applied in a pragmatic manner. Those tensions have also been observed in studies and other evaluations, and are not just present in DC. A thematic evaluation of the visibility of EU external action91 found that “Partnerships with other organisations (UN, governments, NGOs, etc.) are vital in EU external action, but there is a trade-off as visibility needs then to be shared. This remains a source of tension and too much insistence being put on EU visibility can undermine the effectiveness of the cooperation and the sense of ownership felt by partners.” Tensions between visibility and aid effectiveness do not only arise around DC, but also other donors face these tensions. This was reflected in the problematic division of labour discussions in many case study countries, and the fact that many donors want to keep their ‘seat at the table’. Finally, in addition to underscoring the practical steps taken to ensure visibility (through communication and visibility activities), the case studies have also pointed out that visibility (for the EU and the DA partner) was also generated by the process of DA itself. As DAs allowed the EU to be often more pro-active in donor coordination together with the implementing partners, this did contribute to the visibility of the EU and the MS. This also reflected the fact that the EU was rarely a silent partner when delegating. This suggests that, independently of whether visibility activities were or not fully implemented, the EU (and the DA partners) appears to have rarely ‘lost’ visibility as a result of DA. Findings from the questionnaire These main findings from the case studies are broadly in line with the responses to the questionnaire, as indicated in the table below. Respondents were largely of the view that concrete actions have been taken to ensure visibility of the EU in the case of DAs. 74% of respondents either agreed or strongly agreed. Secondly, most respondents also agreed that the EU visibility was generally satisfactory in DA-related projects (9% strongly agreed, 57% agreed) while a number also disagreed (26%).

91

Consortium PARTICIP-ADE-DIE-DRN-ECDPM-ODI, Thematic Evaluation of the Visibility of EU external action 2005-

2010, June 2012.

78


Figure 3.6.1 EUD questionnaire response: Concrete actions have been taken to ensure visibility of the EU in the case of DAs


0% 20% 40% 60% 80%

%

1 Strongly agree 4 11

2 Agree 22 63

3 Disagree 6 17

4 Strongly disagree 0 0

5 No opinion 3 9


Figure 3.6.2 EUD questionnaire response: In case of DAs, the visibility of the EU was satisfactory


0% 20% 40% 60% 80%

%


2 Agree 20 57

3 Disagree 9 26


5 No opinion 3 9


3.6.5 Visibility of the delegating donors in Transfer Agreements (JC 6.2)

JC 6.2. Concrete actions have been taken to ensure the visibility of the delegating donors in case of Transfer Agreements.

Introduction The template for TA Assessment Fiches92 includes one visibility question, namely: “what will be the additional impact on European visibility with this delegated cooperation?” It is interesting to note that the latter question refers to the broader concept of ‘European visibility” while the two questions concerning the DAs referred specifically to the EU visibility, i.e. of the delegating partner. Article 9 of the General Conditions attached to a TA refers to visibility in the following way:

“Unless the donor requests or agrees otherwise, the Commission shall ensure the visibility of the donor's contribution”;

92

Both the version of 2012 and the version used prior to 2012.

79


“The donor accepts that the Commission publishes in any form and medium, including on its website, the name and address of the donor, the purpose of the contribution as well as the amount contributed and, if relevant, the percentage of co-financing.”

This article is much less detailed than the corresponding article of the General Conditions of DAs (see previous section). In other words, the visibility requirements mentioned in the TA assessment fiches are less strong and detailed than in case of DAs. Findings from case studies In the following table the scores regarding 31 TAs in 13 projects and programmes are summarized, which shows that in nearly all TA projects (11/13), some actions were taken or partly taken to ensure visibility of the TA partners. With the majority of projects showing a score of ‘partly’, the overall picture vis-à-vis visibility for TAs is however less convincing than for DAs. Table 3.6.2 Case study scores: Extent to which actions have been taken to ensure the visibility of the EU and the TA partners

Country Yes Partly No Total

Benin 1 1 0 2

Ghana 0 0 1 1

Mali 1 0 0 1

Mozambique 0 3 0 3

Nicaragua 0 0 0 0

Palestine 1 1 1 3

Tanzania 0 1 0 1

Timor-Leste 0 0 1 1

Haiti 0 1 0 1


Evidence from case studies suggests that some action has been taken to ensure visibility of TA partners, but that such action has not been as systematic and comprehensive as in DA related projects and programmes, partially reflecting the lack of clear specification on this in TAs. In most cases, the EU mentioned the logo of TA partners in project documentation and specified who is funding the project or activity. TA partners were also invited to key projects events. There is no evidence of any communication/visibility plans having been prepared in TA projects though, as this was not a requirement. In a few cases, implementation of visibility requirements was weak in the early phase of the TA project, before it then improved, following pressure from TA partners. This was the case of the APONTRA in Haiti where the EU had made little reference to AFD in its reports, but that was (partially) addressed following AFD request. When visibility was relatively low for TA partners, this was in general less of a concern for the TA partner. This partly reflected the nature of the delegation. The TA partners delegating funds to the EU were in general ‘silent partners’. In the case of PEGASE in Palestine, for instance, visibility was low for the TA partners (which were contributing direct financial support to PEGASE recurrent expenditures), but that was not a major concern to them. On the side of the EU meanwhile, the nature of the TA relationship also meant that there was less need to address visibility concerns. In comparison to DA projects where the EU was a large contributor, the TA partners were contributing less proportionally to the TA-related projects and programmes. At the same time, the TA partners’ funds were never

80


earmarked (contrary to DA funds) or consisted of budget support, meaning that there was no need to report separately on the funding of the TA partner’s contributions. Finally, visibility of the TAs related to GCCA was generally found to be low. This partially reflected the small contributions of the TAs to the GCCA actions. Findings from the questionnaire The main findings from the case studies are mainly in line with the responses to the questionnaire, as indicated in the table below. Respondents were broadly of the view that concrete actions have been taken to ensure visibility of the TA partners in the case of TAs. While the majority of respondents (60%) had no opinion on the issue, 34% were of the opinion that concrete actions have been taken to ensure the visibility of the TA partner in the case of TAs (against 6% who disagreed). Secondly, more respondents agreed (26%) or strongly agreed (6%) that the visibility of the TA partner was generally satisfactory in case of TAs than disagreed (9%). Such results have to be interpreted with caution though, given the limited number of respondents and the possible bias in the perception of EU officials on this issue. When compared with the results in relation to DAs, those findings underscore, as the case studies do, that the extent to which actions were taken to ensure visibility in the case of TAs was in general lower than in the case of DAs, and that overall visibility of TA partners was more limited than visibility of the EU in the case of DAs. Figure 3.6.3 EUD questionnaire response: Concrete actions have been taken to ensure visibility of the DC partner in the case of TAs


0% 20% 40% 60% 80%

%


2 Agree 12 34

3 Disagree 2 6


5 No opinion 21 60


Figure 3.6.4 EUD questionnaire response: In case of TAs, the visibility of the DC partner was satisfactory


0% 20% 40% 60% 80%

%


2 Agree 9 26

3 Disagree 3 9


5 No opinion 21 60


81


3.7 EQ 7: Balance of DAs and TAs

EQ 7: What have been the main reasons why to date, the number and value of TAs have been much lower that the number and value of DAs?


There is a considerable imbalance between the number and value of TAs and those of DAs. Since the introduction of DC, 164 DAs have been signed with a value of € 1,263 million and 59 TAs with a total value of € 291 million. This EQ addresses the causes of this imbalance by considering the constraints of engaging in TAs and the relative advantages of concluding a DA. Different types of constraints have been analysed (amongst others administrative, procedural, managerial, budget-related, implementation-related) on the side of both the EU and the DC partners with regard to TAs, and several hypotheses have been tested as regards the advantages of DAs vis-à-vis TAs, in order to explain the above mentioned imbalance.


In an internal Guidance Note issued in December 2007, the Commission introduced two new aid “implementation modalities”: DAs and TAs93. In case of DAs, the Commission delegates authority to manage EU funds to another donor, and in case of TAs another donor entrusts the management of its funds to the EU. That Guidance referred also to the balance between the two instruments by stating that DC should not be a one way process (mainly DAs) and reciprocity should be envisaged (but not on a case by case basis). In 2009, a note was circulated to the EU DG AIDCO Directors and EU Heads of Delegations dealing with establishing an “adequate balance” between DAs and TAs. More precisely, the following targets were set:

“At global level, the total amount of TAs shall be in principle at least half of the total amount of DAs”;

“At the level of every relevant entity the total amount of TAs from a delegated entity or the relevant Member State shall be in principle at least half of the total amount of DAs with the same entity”. This would be calculated only on the basis of a minimum of five operations94.

In other words, the ratio between TAs and DAs should be at least 50% both at the global level and at the level of each partner.

The report on Delegated Cooperation (2007-2012), which served as an input for the revision of the DC guidelines in 2012, concluded that the criteria were not met in the period under review. In fact up to early 2012, a total of 94 DAs had been signed or were being prepared with a total value of €489 million, against only 36 TAs with a total value of €159 million. Moreover, a negative trend was observed, because the number and amount of DAs was increasing much faster than those of TAs. At the level of individual entities, only two organisations met the 50% ratio. The report concluded that so far the balance

93

“Guidance on joint co-financing with Member States and other bilateral donor”, AIDCO D(@007) 24585. 94

Note to AIDCO Directors and to Heads of Delegation, Delegated Cooperation – framework, Ares (2009)294406 –

26/20/2009, (internal document).

82


criteria had been indicative, but that a more strict application would be needed in order to meet the targets of these criteria95. However, mid 2012 in the process of drafting a new DC Guidance Paper, the Commission decided to remove the two indicative TA/DA balance criteria. The focus on the balance was dropped to the benefit of a “value added criteria based on a cost/benefit analysis”96. This was confirmed in the new Guidance Paper issued in July 2012 stating that “there are no formal balance criteria applicable …. but Member States are strongly encouraged to delegate funds to the Commission (in particular for Budget Support operations) and participate in other innovative forms of cooperation with the Commission, such as the EU trust funds, in order to maintain a global balanced approach towards EU cooperation”97. This view was still relevant in 2014, as the information note on DC published in 2014 reads: “A balanced approach is expected but no explicit ratio is set. This means that over time MS bodies/agencies should also delegate funds to the Commission through Transfer Agreements”98. Although establishing an “adequate TA/DA balance” is no longer an objective (a criterion) of the Commission’s DC strategy, it has nevertheless been decided to devote one EQ to it, because that criterion has played an important role up to 2012. Moreover analysing this issue may deepen the understanding of the pros and cons of both TAs and DAs.


Since 2012, the concept of reciprocity has been broadened: it does not only concern the TA/DA balance, but also other ‘innovative’ ways of cooperating with the Commission, such as EU Trust Funds, should be taken into account. This broader view on reciprocity was supported by various staff members from the EU and DC partners interviewed by the evaluation team. However, it should be noted that that broader concept of reciprocity is not part of the scope of this EQ. It should also be noted that this EQ is not directly related to other outputs or outcomes of the Intended Effects Diagram. An adequate balance between TAs and DAs is for instance not a prerequisite for an optimal division of labour or improved aid effectiveness. As a consequence, this EQ should be seen as a rather operational question focussed on assessing why the TA instrument has been less attractive than the DA instrument, with a special focus on assessing the constraints of preparing and signing (more) TAs.

3.7.4 Inventory of TA/DA ratios

Before answering the question why the DA instrument has attracted much more interest than the TA instrument, it is useful to get first a clear picture of the actual balance between TAs and DAs. The development of the global TA/DA ratio has been analysed in section 2.4 and showed a growing imbalance between TAs and DAs over time, which was not only caused by a stagnating number of TAs signed per year, but also by a substantial increase of the number of DAs signed per year. There are big differences in the TA/DA ratios across EU-MS (see table 3.7.2). The two MS making most use of the DA instrument (Germany and France) have concluded only a few

95

Report on Delegated Cooperation 2007-2012, Ref. Ares (2012)388917 – 30/03/2012 (internal document). 96

Note to the File “Revision of the document "Guidance on joint co-financing with Member States" - Comments received

from services”, July 2012 (internal note EC). 97

Guidance paper on Delegated Cooperation, 2012, REF. ARES (2012).916012 – 27/07/2012. 98

Information Note on Delegated Cooperation (CODEV 11/9/2014).

83


TAs and have thus a very low TA/DA ratio (respectively 1% and 16%). As these two countries make up for about 65% of the number of DAs (and 70% in value terms), their low TA/DA ratios have a big impact on the overall TA/DA ratio. Most other MS have also a TA/DA ratio below the indicative target of 50%: Portugal (0%), Italy (3%), Spain (13%), Denmark (18%) and the Netherlands (28%; all these percentages refer to the amounts in millions of Euros.) Relatively smaller MS such as Belgium and Austria have established an adequate balance between TAs and DAs. Other MS such as Ireland and Sweden have only signed TAs and no DAs at all. Table 3.7.1 Portfolio analysis: TA/DA ratios per DC partner

DC partner Number of contracts Amounts (millions of €)

DA TA % TA of DA DA TA % TA of DA

Australia 1 1 100 4.0 1.8 100

Austria (ADA) 3 7 233 10.0 7.5 75

Belgium (BTC) 8 9 113 25.3 47.8 189

Canada - 1 ∞ - 2.2 ∞

Czech Republic - 1 ∞ - 0.2 ∞

Cyprus - 2 ∞ - 1.2 ∞

Denmark 7 4 57 57.8 10.2 18

EIB - 1 ∞ - 48.3 ∞

Estonia - 1 ∞ - 0.8 ∞

France (AFD+ FEI+ADETEF) 35 2 6 310.9 49.1 16

Germany (GIZ+KWF) 72 1 1 574.3 3.5 1

Ireland - 6 ∞ - 30.4 ∞

Italy 3 1 33 34.9 1.1 3

Japan - 2 ∞ - 6.2 ∞

Luxembourg (Lux-dev) 2 5 250 10.4 5.7 55

Netherlands 1 2 200 19.8 5.5 28

Netherlands Antilles 3 - 0 42.6 - 0

Portugal (CICL) 7 - 0 29.2 - 0

Spain (AECID + FIIAPP) 11 2 18 92.8 12.5 13

Sweden - 6 ∞ - 25.8 ∞

Switzerland - 1 ∞ - 2.3 ∞

UK (DFID + British Council) 11 4 36 51.5 29.1 57

Total 164 59 36% 1.263.4 291.2 23%

It is also interesting to have a look at the TA/DA ratio per EUD (although that was not the focus of the TA/DA balance criterion). Out of the 61 EUDs having concluded at least one DC agreement99, 35 signed only one or more DAs, 20 EUDs signed both DAs and TAs and six EUDs signed only TAs. Out of the 20 EUDs having both TAs and DAs, only four of them had a TA/DA ratio higher than 50% (in value terms). Looking specifically at the EUDs/countries involved in the case studies of this evaluation, their TA/DA ratio is actually not far from the original indicative target of 50%. Only Nicaragua (no TAs) and Timor-Leste (17%) have a TA/DA ratio of less than 50% (expressed in number of agreements). In money terms the TA/DA ratios are lower. That ratio does not reach 50% in six of the nine case study countries, while three case study countries have a ratio well above 50% (In Ghana and Haiti, the volume of TAs is even higher than the volume of DAs). Finally it is worth noting that 12 of the 59 TAs were related to the GCCA programme100 managed by the EU HQs and 13 TAs were supporting the PEGASE Direct Financial Support system in Palestine.

99

Some DAs and TAs are managed by HQ instead of a Delegation. 100

Formally there were only 7 TAs supporting 12 GCCA projects, but in the Commission’s database they have been

counted as 12 TAs.

84


3.7.5 Constraints at the level of the EU (JC 7.1)

JC 7.1. The number and value of TAs were relatively low because of constraints at the level of the EU.

This JC considers constraints at the EU HQ and EU Delegations which may have affected negatively the number of TAs having been signed so far. The following possible constraints have been explored: administrative and procedural constraints, management constraints at the level of the EUDs, mismatch between the EU’s focal sectors and the sectors of interest of the MS, and inability of the EU to promote the added value of the TA modality. The procedures for preparing and concluding a TA are less labour intensive than those of a DA. The TA partner does not have to go through a Pillar Assessment and the proposal for a TA does not have to be assessed by the Commission’s Quality Support Group. The procedures fixed in 2007 stated that AIDCO had to draft a proposal for accepting additional funds via a TA or – in case of co-financing – had to include such a proposal in the Financing Decision, in both cases to be approved by the College of Commissioners or the Commissioner in charge101. After approval, the TA document could be signed by the EU Commissioner for Development Cooperation. In 2013, decision-making was decentralised to the Commission’s Directorates. In 2012 a new TA Assessment Sheet was introduced, which had to be approved by the Director of the relevant Directorate. Subsequently, that Assessment Sheet had to be further completed and attached to the Action Fiche of the Financing Decision concerned and included in the Annual Action Programme (AAP). It was suggested that the TA partner writes a letter of intent, which should then be attached to the Action Fiche as well. The actual TA document could only be signed after the AAP was approved102. However, very few TA Assessment Fiches have been uploaded in the EU information system (CRIS) and these documents could also not be traced during the missions carried out in the case study countries. The transfer of decision-making (as regards concluding a TA) from the Commissioner to the Directorate level in 2013 has facilitated the TA preparation process within the EU103. Three constraints in the procedures before signing a TA have been identified. First, it was mentioned during the interviews held with staff of the Commission, the EUDs and the DC partners, that there was confusion about the precise content of the procedures. This is also illustrated by documentation available in CRIS, which includes e-mail correspondence between different DEVCO units on clarifying procedural matters. Second, many actors were involved in the TA preparation process: the EUD, EU Headquarters, the Headquarters of the DC partner and sometimes their development cooperation agencies, and the partner country government. Third, the different procedures and decision-making cycles of the EU and the donors caused delays in the process. At the EU side, a TA can

101

Note for the attention of the Heads of Delegation, AIDCO D(2007) 24585, December 2007. Internal document. 102

European Commission Directorate General for International Development and Cooperation, DEVCO Companion to

Financial and Contractual Procedures applicable to development and cooperation financed from the general budget of

the EU and from the 11th EDF, Version 4.1, September 2014. 103

However, this de-concentration of decision-making has made it more difficult to monitor properly the implementation of

the DC policy, including the TA-DA balance (which is however no longer a strategic issue since 2012). Initially

development and monitoring of the implementation of the DC policy as well as decision making as regards individual

DC agreements, was centralised in Unit O2 of DEVCO, which had thus the overview of policy

application/implementation, the trend of the total numbers and value of the DC agreements and the TA/DA balance. In

March 2013 it was decided to de-concentrate the decision making to the geographic directors and to charge Unit A2

with the development and monitoring of the DC policy and Unit R1 with monitoring the number and amounts of DC

agreements.

85


only be signed after the Financing Decision of the TA-funded project has been approved (or amended in case of an on-going project). On the other hand an agreement in principle to conclude a TA should have been reached much earlier, because the TA has to be taken into account in the Financing Decision of the Commission and in the relevant Action Fiche. In a presentation of DEVCO on the State of Play of Delegated Cooperation (2011), the need to anticipate a Transfer Agreement at an early stage, was cited as main the “constraint”104. EU Member States and MS agencies have their own approval procedures and processes, which are not per se aligned with those of the EU. Such a mismatch may cause considerable delay. For instance, in the case of four of the 31 TAs being part of the nine country case studies, there was a delay of more than six months between the date of signature of the TA by the TA Partner and by the EU. As regards the possible inability of the EU to promote the EU’s added value for managing TAs, it has to be mentioned that it was not so much an inability as well as a lack of urgency after the Commission had decided to drop the 50% target as regards the TA/DA ratio in 2012. During the last couple of years, new modalities as regards cooperating with the MS have been developed, such as the Trust Funds, blending of financial instruments and Joint Programming, which appear to be (more) effective instruments to mobilise additional funds from the MS. These instruments have also other added values, such as mobilising loans from financial institutions and banks through blending and stronger coordination and more co-financing via EU Trust Funds.

3.7.6 Constraints at the level of the (potential) TA partners (JC 7.2)

JC 7.2. The number and value of TAs were relatively low because of constraints at the level of the transferring partner.

This JC investigates what obstacles are at play at the level of the DC partner / Member State when considering a Transfer Agreement. These include constraints of administrative and procedural nature, budget constraints, the risk of losing visibility, etc. For most donors, the decision to conclude a TA is taken at the level of their Ministry for Development Cooperation, and not at the level of their representation in the partner country or an agency responsible for the implementation of development cooperation (GIZ, KfW, AFD, BTC, etc.). This implies that the official procedures and processes of ministerial approval have to be respected. When these are not fine-tuned with the EU decision making process, the preparation of a TA can take a long time, which could be considered as a hindrance for setting up a TA. Some MS are reluctant to transfer additional money to the EU on top of their regular contributions to the EU budget and the EDF. This opinion could be considered as the major hindrance for expanding the number and amount of TAs. These MS argue also that reciprocity should be considered in a broader context, including Trust Funds and joint and parallel co-financing of projects organised at the level of their representative offices in the partner country (without using the TA instrument). In recent years, budget constraints are also (increasingly) playing a role and demotivate MS to sign (more) TAs105. DC partners need their entire aid budget for financing their

104

Lionel, Atlan, EuropeAid/02, Delegated cooperation: State of Play (PPt presentation, 13 December 2011). 105

This was much less the case in the early years of the evaluation period. Moreover, several TAs have been signed in

November and December, which indicates that TAs were also used as an instrument to quickly spend unused funds

86


projects and programmes in their focal sectors and priority countries. Moreover, in particular the larger MS feel that they have sufficient capacity to implement their own shrinking budget. Loss of political and strategic influence and loss of visibility are also factors explaining the low interest of MS to conclude TAs; especially at the level of the larger MS. Complex and time consuming financial EU procedures have discouraged MS to get involved in the preparation of more TAs. The use of EU procedures has also been a reason for some MS to abstain from signing a TA, because they saw the EU rules as less effective or flexible compared to their own procedures. Moreover, the contractual obligations of the EU towards the TA partner are very lean – in particular compared to the contractual obligations of a DA partner – which has also discouraged potential TA partners. The TA portfolio reveals that most of the TAs were concluded with the smaller MS. Three of the explanatory factors are: (i) smaller MS do not have sufficient capacity to manage themselves the implementation of their entire development cooperation budget, (ii) smaller MS have embassies in fewer countries and a TA offers the opportunity to finance a project in such a country, and (iii) smaller MS are less concerned about the loss of visibility when spending their money via TAs. The opposite of these factors are some of the reasons why the larger MS are not really interested in concluding more TAs. The TA/DA ratio has never been a strategic issue for the MS. They also did not have a steering mechanism to influence that ratio. That would also have been difficult – although not impossible – because the preparation and formulation - and sometimes even the decision making – is often done by different entities. This is particularly the case when a MS has a Development Cooperation Agency being (partly) independent from the Ministry responsible for development cooperation.

3.7.7 DAs are much more attractive than TAs (JC 7.3)

JC 7.3. DAs are much more attractive than TAs.

The (growing) imbalance between TAs and DAs was not only caused by a stagnating number of TAs signed per year, but also by a substantial increase of the number of DAs signed per year. This JC is therefore focused on analysing the advantages or attractiveness of DAs compared to TAs. Four important factors promoting the use of the DA modality and discouraging the use of the TA modality are:

DAs reduce the workload of the EUDs - while TAs will increase their workload – which is an important consideration for EUDs to invest time in preparing DAs (rather than in preparing TAs), in view of their capacity constraints. It is worth mentioning that also the internal review of Delegated Cooperation 2007-2012 concluded that reducing their workload was for many EUDs the main reason to conclude DAs;

DA partners are interested to sign more DAs because DAs are a source of additional income for them, which allows them to expand the scope and size of their interventions and strengthens their Development Departments (and/or Development Agencies);

towards the end of the fiscal year. From the 23 TAs of the case studies for which a signature date was available, 4

were signed by the TA partner in November and 7 in December.

87


DAs provide good opportunities to implementing agencies to mobilise additional funds without becoming involved in competitive procedures (open or restricted tenders; call for proposals). This stimulates the use of the DA modality;

Large MS have sufficient implementation capacity to spend their entire budget and manage all the projects (thus no interest in TAs), whereas smaller MS have less capacity to spend their budgets and are therefore more interested in transferring funds to the EU.

In the questionnaire filled in by 35 EUDs, the EUDs were asked to give their opinion about five possible reasons why there are more DAs than TAs. The answers to that question are summarised in table 3.7.3 hereunder. The most remarkable feature of that table is the high level of scores for “no opinion” illustrating that the TA/DA ration was not a (strategic) issues the EUD was paying attention to, and possibly because EUDs felt they did not have sufficient view on DC strategy issues to allow then to answer the question. Other features of table 3.7.3 are:

Most EUDs having expressed an opinion, disagreed with the statement that “the EU was faced with more constraints to sign a TA than a DA”, and also with the statement that “the DC partners were faced with more constraints to sign a TA than a DA”;

Most EUDs having expressed an opinion, agreed with the statement that “DAs are much more attractive to DC partners, because they increase their scope of activities”.

EUDs having expressed an opinion had contrasting opinions about the statement that “EUDs aim for reducing their work load: Thus more DAs than TAs” and the statement that “the EU has sufficient funds available for DAs, while DC partners are faced with a scarcity of funds making it difficult to finance more TAs”.

Table 3.7.2 Case study scores: Reasons why there are more DAs than TAs

Reason Strongly

agree Agree Disagree

Strongly

disagree

No

opinion

EU faced with more constraints to sign a TA than a DA

1 3 11 0 20

DC partner faced with more constraints to sign a TA than a DA

1 2 9 0 23

DAs much more attractive to DC partners, because they increase scope of activities

5 16 3 0 11

EUDs aim for reducing their work load: Thus more DAs than TAs.

5 9 8 3 10

The EU has sufficient funds available for DAs, while DC partners are faced with a scarcity of funds making it difficult to finance more TAs.

4 9 8 1 13

Source: Questionnaire filled in by 35 EUDs.

Finally, the special Palestinian case is worth mentioning here. The TA/DA ratio is very high in Palestine: about 1.6 in number of agreements (15 TAs and 9 DAs) and 0.64 in value terms (€ 35 million TA value and € 55 million DA value). The high number of TAs is explained by the fact that the EU has a clear comparative advantage to lead the politically sensitive PEGASE Direct Financial Support (DFS) mechanism. Instead of funding PEGASE DFS directly, a number of MS prefer to channel the funds via the EU using the TA modality, in order to share the political risk of that kind of support and to make use of the comparative advantages of the EU.

88


3.8 EQ 8: Assessment of DC proposals

EQ 8: What has been the quality of the decision making process and the assessment of DC proposals in view of the DC objectives and assessment criteria as defined by the EU?


This EQ focuses on assessing the quality of the preparation of a DC proposal, with a special focus on (i) the motivation of EU Delegations to start preparing a DC proposal and (ii) the content and quality of the DC Assessment Fiches and related Action Fiches of the projects which will be funded. The scope of this EQ consists of the DAs and TAs of the selected case studies. The EU is supposed to use standard fiches (one for DAs and one for TAs) for assessing DC proposals. This requirement is explicit in the cases of DAs, where the Assessment Fiche is attached to the Action Fiche being used for taking a Financing Decision regarding the proposed DA. In case of a TA, the Assessment Fiche is used for taking a decision on whether the transfer from another donor will be accepted. These fiches are potentially one important tool for translating the DC policy and strategy into practice. This EQ is focused on analysing whether the fiches provide sufficient information for taking well-founded decisions as regards approving or disapproving DC proposals by looking at both the appropriateness of the format and the quality of the completed fiches. Furthermore, the actual motivations of Delegations to enter into a DC agreement are also taken into account and are compared with the official DC policies and guidelines.


The rationale for this EQ relates to the fact that the assessment whether or not DC project/programme have achieved its objectives and is in line with the EU’s DC policy and strategy, as analysed under EQs 1 to 5, partly depends on the quality of the decision-making process leading to DC, including the use and the quality of the Assessment and Action Fiches. Over the years 2007-2012 several Notes and Guidance Papers have been issued stipulating the general criteria DC proposals have to comply with. This culminated in the following list of 10 assessment criteria established in 2012106:

The programme is owned and led by the partner country and two or more donors, including the Commission, have agreed to fund;

The partner country agrees with the Delegated Cooperation arrangement;

The fund managing donor has adequate technical and financial management capacity;

The fund managing donor will remain active in the programme/sector in the near future;

Delegated Cooperation results in visibility gains for the EU and efficiency gains and reduced transaction cost for both the EU and the partner country;

Delegated Cooperation results in larger and cost effective programmes with a measurable/quantifiable operational impact;

106

Guidance paper on Delegated Cooperation with Member States, 2012.

89


Delegated Cooperation covers large programmes and the delegated EC amount is not less than €3 million;

Appropriate visibility of each of the donors involved is ensured;

Pure technical assistance projects should not be considered for DAs;

Co-financing should, in principle, be a major prerequisite for DAs. It is expected that these criteria are reviewed in the DA and TA Assessment Fiches and the Action Fiches of the projects concerned. The specific aim of this EQ is to consider whether those criteria have actually been assessed and taken into account when taking financing decisions as regards the DC proposals, and as such review the quality of the DA assessment process. Although the above mentioned list was established in 2012, it was/is in fact a formalisation and consolidation of criteria most of which were already in use since 2008. It is therefore justified to use this list also for evaluating DC proposals dating from the period prior to 2012.


The Evaluation Team has faced some methodological challenges in addressing this question. One weakness is the fact that DC assessment fiches, in particular TA assessment fiches, are missing. In the 33 DA projects analysed, only 13 DA fiches were collected out of the 44 DAs assessed. Meanwhile, only one TA fiche was found (in Mali) for the 40 TAs assessed. In the case of DA assessment fiches, this was mostly fiches for DA projects signed prior to when the requirement was made that such fiches be filled (in September 2012). The way around this was to ask for the rationale of the decision-making in interviews and our assessment is based on a reconstruction of the decision-making process. In some cases, it was difficult to find people that were involved in the decision-making process at the time, but we could still reconstruct it to the extent possible. In some other cases, making that reconstruction was particularly challenging. Given the availability of one singe TA fiche (for the PAOSC project in Mali), a specific issue we have faced is that we cannot assess the quality of TA fiches on the basis of one example.

3.8.4 Motivation of EUD to enter into DC agreements (JC 8.1)

JC 8.1. The motivation of the EU Delegations to enter into a DC agreement is in line with the objectives of DC as defined in the EC’s Guidance Paper.

Introduction This JC is focussed on analysing the motivations presented by EUDs to DG DEVCO to enter into a DC agreement. Normal procedures of Project Cycle Management have to be followed when preparing and formulating the project (Action in EU jargon), including drafting an Identification Fiche and later on an Action Fiche, while specific DC Assessment Fiches have to be filled in for explaining and justifying the DC funding. Action Fiches have to be reviewed by the DG DEVCO Quality Support Group 2. The DC Assessment Fiche should be the key document that “includes an assessment of the political relevance and feasibility of the project/programme for the EU, for the partner country and for the Delegatee body”. It asks whether the relevant delegated cooperation contributes to the strengthening/rationalisation of a wider division of labour process, whether the delegated action is coherent with EU strategies and programming documents, its impact as regards EU visibility, cost-benefit/impact ratio and the likely risks and

90


disadvantages of the delegated action. With regard to the partner country it asks whether the delegated action is aligned to its development strategy, the action is owned and led by the partner country and whether it agrees with the mode of management of the action. It also asks about the expected efficiency gains and reduction of transaction cost of the action. As regards the Delegate body, the Assessment Fiche asks whether it has the technical and financial management capacity to manage donor funds and implement the action (e.g. international track record, sectors of intervention, investment pipeline), whether it provides comparative advantage and whether it has significant and sustainable involvement and investment in the relevant programme/sector”107. Delegations are requested to prepare DC Assessment Fiches. The templates were revised in July 2012 and were presented as an Annex to the new DC Guidance Paper also issued in 2012. Prior to 2012, the requirement of preparing Assessment Fiches was much less stringent and the rationale for proposing a DC intervention was often elaborated in the Action Fiche of the project concerned or in an Explanatory Note made by the EUD and addressed to DG DEVCO. Given that only 13 DA fiches (and one TA fiche) could be collected, assessing the motivations of EU Delegations in using DC has been challenging. However, next to the DC Assessment Fiches, existing or not, the Identification and Action Fiches of the DC-supported projects and programmes may contain useful information about why the EUD has opted for using the DC implementation modality. Findings from case studies The case studies have underscored that the use of DA (and TA) assessment fiches has not been systematic. As a result, making a structured assessment of the motivations of EUDs to enter DC agreements, in particular TAs, is challenging. In the 33 DA projects analyzed, only 13 DA fiches were collected out of the 44 DAs assessed. Meanwhile, only one TA fiche was found (in Mali) for the 40 TAs assessed. Seven of these fiches were collected as part of the inventory of CRIS documentation undertaken during the desk phase. During the field phase, the evaluation team asked explicitly for missing fiches to be provided and it received six additional ones. On the other hand, the use of DA assessment fiches has significantly improved since the requirement was made in September 2012 to submit them as annexes to the Action Fiches. Out of the 13 DA assessment fiches retrieved, 11 were for DA signed in 2013 and 2014. This means that out of the 20 DAs of the case study countries signed during those years, we could collect DA fiches for 55% of them. On the other hand, out of the 24 DAs signed prior to 2013, only 2 were collected (8%). On that basis, the discrepancy between on one hand the number of DA assessment fiches found and on the other the number of TA assessment fiches collected can be partly explained by the fact that only three TAs were signed after 2012. Identification Fiches and Action Fiches have also provided some information about why the EUD has opted for using the DC implementation modality. The Action Fiches generally included a section outlining the rationale behind delegating funds to the DC partner (i.e. typically the comparative advantage of the latter), but the information provided was not very comprehensive and the rational rarely elaborated upon, especially if the DA was only one component of a broader EU action (there was rarely more than a paragraph outlining this rationale).

107

Quote from the Guidance paper on Delegated Cooperation with Member States, 2012, p. 8.

91


Notwithstanding these limitations, the case studies provided information about the actual motivations of Delegations to enter into a DC agreement, which suggests that those are broadly consistent with the objectives and the 10 criteria presented in the DC Guidance of 2012. More specifically, interviews with DA partners as well as analysis of the few available DA fiches point out that in most cases DC was mainly motivated by the fact that a specific DC Agency has the required expertise to implement a certain Action. The DA partners (managing the DA funds) have usually a strong expertise and well-established position in the sector or sub-sector concerned, and have often also invested in the design of new projects or programmes, of which the DA-funded activity is part of. Improving comparative advantages and division of labour within a sector has been as such the main motivation for the use of DC, which is consistent with one of the initial envisaged outputs of DC (this point is further elaborated in EQ1). While this was rarely explicitly mentioned in DA assessment fiches, interviews with EU Delegations have also underscored that decreasing the workload of EUDs was often an important consideration for preparing and signing a DA. (Such rationale can be linked to the following criteria in the Guidance Paper of 2012: ‘Delegated Cooperation results in larger and cost effective programmes with a measurable/quantifiable operational impact’). For an EUD, a DA is a relatively simple procedure for outsourcing the implementation of an activity, while the workload for monitoring and control of project implementation is relatively limited (compared to supervising a service contract). In the context of the staff reductions experienced in a number of EU Delegations in recent years, this has been emphasised as an important operational rationale for the use of DC in a number of case study countries, in particular Benin, Mali and Haiti. Findings from the questionnaire These main findings from the case studies are mainly in line with the responses to the questionnaire submitted to EU Delegations as indicated in the table below. In the questionnaire, the EUDs have been asked to assess whether the objectives of the signed DC agreements are well in line with the envisaged objectives as specified in the DC Guidance Notes and instructions. Among 35 respondents, 3% strongly agreed, 46% agreed, 3% disagreed, 3% strongly disagreed and 46% were of no opinion. It should be noted that almost half of the respondents had no opinion. The comments suggest this has to do with the lack of clear guidelines stipulating the main objectives of DC. One respondent noted: “There are no comprehensive DA guidelines nor a comprehensive compilation of guidance notes. At times, different guidance notes appear to contradict each other. There is no complete set of templates to be used for the different steps of preparing and implementing a DA (budget table; reports...).” Another stated: “These Delegation Agreements were drawn up many years ago, since then we have not had new ones and we therefore did not work with the guidelines concerned.”

92


Figure 3.8.1 EUD questionnaire response: The objectives of the signed DC agreements are well in line with the envisaged objectives as specified in the DC Guidance Notes and instructions


0% 20% 40% 60% 80%

%

Strongly agree 1 3

Agree 16 46

Disagree 1 3


No opinion 16 46


More significantly perhaps, as part of the same questionnaire, delegations were also requested to assess a number of potential objectives behind the use of DC as either their ‘main objective’ or a ‘secondary objective’. The EUD respondents to the questionnaire mentioned three main motivations to use DAs and TAs:

1. Use of comparative advantages (76% main motivation and 18% secondary

motivation); 2. Larger projects and programmes (45% main motivation and 30% secondary

motivation);

3. Strengthening the relations between the EU and its Member States (30% main motivation and 58% secondary motivation).

The importance of use of comparative advantages and creating larger projects and programmes is fully consistent with the DC criteria set, and with the findings related to outputs as presented in the earlier sections of this report. The importance given by EUDs to strengthen the EU – Member State relations is in line with the broader context in which DC was developed, but is important to note as it is not a specific objective of DC as mentioned in the DC guidance. While the main motivations for concluding DC agreements indicated in the responses to the questionnaire were broadly confirmed by the case studies, the evaluation team found that the reduction of the workload of the EUDs, which was not considered as a main motivation in the questionnaire, was among the main motivations in many case study countries as underscored above.

93


Figure 3.8.2 EUD questionnaire response: Motivations for using the DA modality (% of respondents assessing the criteria as a main motivation, secondary motivation or as not really an issue)*

* Number of respondents: 33

3.8.5 Appropriate format of Assessment Fiches (JC 8.2)

JC 8.2. The format of the Assessment Fiches is appropriate for making an informed decision

This JC is focussed on assessing the format of the Assessment Fiches so that it contains the required information and adequate analyses, allowing decision makers in Brussels to make informed decisions. The DA Assessment Fiche (version of 2012) is constructed as follows:

context information, including short description of the programme, overall EU contribution, the DA partner involved (Delegatee Body), amount delegated, amount co-financed by the DA partner;

political relevance and feasibility for the EC, in particular whether the delegated action contributes to strengthening/rationalisation of a wider division of labour process, whether the delegated action is coherent with EU strategies and programming documents and whether the action is not a substitute to setting up technical assistance via centralised or decentralised management;

policy relevance and feasibility for the partner country, in particular whether the action is aligned with the partner country's development strategy and is in response to an expressed need and whether the delegated action is owned and led by the partner country and whether it agrees with management of the action by the DA partner;

technical and financial management capacity of the DA partner to manage the project (action) and its comparative advantage to implement the project and whether the DA partner has significant involvement and investment in the relevant programme / sector and whether this will be sustained in the short/medium term.

A separate Assessment Fiche exists for TAs, which is relatively similar to the DA Assessment Fiche but adapted to the context of a TA. For instance specific questions are

0% 20% 40% 60% 80% 100%

Phase out the EU support to a non-focal sector

Improve the division of labour amongdonors within a sector

Clear comparative advantage of theDC partner

Pooling of funds (joint co-financing).

A project or programme with a largerscope and budget

Costs saving by using commonprocedures

Reducing the EUD's workload

Strengthening the relations betweenthe EU and its MS

Preference by the partner country onthe DA modality

Main motivation

Secondary motivation

Not really an issue

Don't know

94


asked as regards the objectives of transferring the funds to the EU and how the delegated action links with more efficient aid modalities. The Assessment Fiches provide a summary overview why an EUD wishes to enter into a DC agreement. A filled-in Fiche should in principle be limited to four pages. Findings from case studies Analysis of the template (version 2012) and of the available DA fiches (prepared after 2012) suggests that the DC Assessment Fiches are broadly copying the objectives as specified in the guidance. However, they clearly suffer from the lack of specificity in the guidance notes. Questions on high-level objectives such as ownership and alignment have been formulated in a very broad manner, and for issues such as transaction costs the Fiche asks to explain what are the ‘quantifiable efficiency gains’. To respond to such questions within a limit of four pages inevitably leads to superficial responses. Furthermore, it is surprising that, while the guidance of 2012 specified that only in certain circumstances non-co-financed DAs are allowed and that the DA contribution should be at a minimum of € 3 million, there is no specific question in the Fiche that requests to elaborate if these criteria are met (from the portfolio analysis, it is clear that after 2012, a number of DAs were signed that were lower than € 3 million and/or not co-financed). In the first version of 2012, no clear connection was made to DC in focal/non-focal sectors. In later versions of the DA Assessment Fiche, if the action was in a focal sector, a ‘thorough justification’ was requested. As a result, the DA Assessment Fiches retrieved provide at best between the lines a justification for the DA – as the responses written to the questions are not really answering the question (as this is often not possible). An example:

“Question: Efficiency gain and reduced transaction cost? Response: There is no alternative to this delegated cooperation as GIZ is the leading organisation in the TVET sector”108.

While the response does provide insight in the motivation of the EU to sign a DA, it says nothing about the size of the transaction costs. It is also worth mentioning that the fact that the DA assessment Fiches have to be approved only at the time of Quality Support Group 2 (with the Action Fiche and the Technical and Administrative Provisions) and not at Quality Support Group I (together with the Identification Fiche) can result in such fiches being viewed as too much of an ex post justification of a decision (to delegate) that has de facto been made at the time of identification. As such, the late stage in the project formulation process in which the DA assessment fiches are reviewed limits their relevance. However, there were also cases where in the Action Fiche, a different modality was foreseen – which was later changed to DC. No accompanying justification notes or Assessment Fiches were found. The format of the TA assessment fiches is broadly similar to the one of the DA assessment fiches. The same conclusions regarding the lack of direct alignment with the criteria as listed in the 2012 Guidance Paper and the lack of specificities in some of the questions asked can be made. Findings from the questionnaire In the questionnaire, the EUDs have been asked to assess whether the DA and TA Assessment Fiches cover all essential issues to be clarified and analysed when taking a

108

DA Assessment Fiche “Occupied Palestinian Territory/ Support to the Private Sector in the West Bank and Gaza Strip –

Component 3 : support to TVET Sector in the Gaza Strip".

95


decision on whether or not to conclude a DC agreement. Among 35 respondents, 6% strongly agreed, 46% agreed, 9% disagreed, 3% strongly disagreed and 40% were of no opinion.

Again, the high number of ‘no opinion’ should be noted: almost half of the respondents have no opinion of or disagree with the quality of the fiches. It was noted by one respondent that “the ever changing rules and templates present a challenge”. Figure 3.8.3 EUD questionnaire response: The DA and TA Assessment Fiches cover all essential issues to be clarified and analysed when taking a decision on whether or not to conclude a DC agreement


0% 20% 40% 60% 80%

%

Strongly agree 2 6

Agree 16 46

Disagree 3 9


No opinion 14 40


3.8.6 Quality of DA Assessment and Action Fiches (JC 8.3)

JC 8.3. The DA Assessment and Action Fiches (actually submitted) provide sufficient and relevant information for taking a deliberate and well-motivated Financing Decision as regards DAs.

The objective of this JC is to assess the quality of information provided in the few available Assessment Fiches and Action Fiches. Findings from case studies As explained above, although the questions in the fiche were not really answered, the few DA Assessment Fiches provided insight in the motivation to opt for a DA. As a result of the lack of direct reference to a number of criteria in the questions in the fiche, a number of the conditions that need to be satisfied for using the DA modality, as laid down by the Commission in its Guidance Paper on Delegated Cooperation, were not systematically met in the available fiches. In the DA with AFD as part of the Support to the Private Sector Project in Benin, the fiche provided information about the rationale for delegating funds to AFD (on the basis mostly of AFD’s comparative advantages), but less was mentioned about whether the DC would contribute to larger and cost-effective programmes, and on co-financing (areas where the DA was relatively weak). Meanwhile, the Identification Fiches, Action Fiches and the Technical and Administrative Provisions attached to the Financing Agreements often included some information about the rationale for going for DC and the choice of the DC partner. The Action Fiche in particular included often included a justification about the DA and the choice of the DA partner but, especially if the DA concerned only one component of a broader EU action, it was relatively limited (no more than one short paragraph).

96


On the basis of the lack of DA Assessment Fiches but taking into account that a justification was generally provided in the Action Fiche, the quality of the decision-making process for most DA proposals among 33 DA projects analysed has been categorised as generally average. Among those projects, 10 were assessed as good, 21 as average and two as weak, as the table below underscores. Table 3.8.1 Case study scores: Quality of the decision making process and assessment of DA proposals

Country Very good Good Average Weak Total

Benin 0 1 4 0 5

Ghana 0 0 2 0 2

Mali 0 0 3 0 3


Nicaragua 0 1 3 0 4

Palestine 0 2 5 2 9

Tanzania 0 1 1 0 2


Haiti 0 0 2 0 0


The ‘average’ score illustrates what was found during the case studies: often no clear written explanation (notably in the form of a DA assessment fiche) was found, but sufficient information was provided to reconstruct the main motivations to engage in a DA. A main shortcoming is the lack of a structured process that is followed when preparing a DA. This is because the DC guidance is not sufficiently clear, but also because the templates are changing frequently. Furthermore, a difference was observed between countries: in Nicaragua, all four DAs had a DA assessment fiche (although for three projects, the decision making was still deemed average, as the documentation failed to explain why the EU was entering new sectors); for Mali, which has eight DAs, not a single DA fiche as found. In Palestine, four fiches were retrieved. Nicaragua and Palestine are funded via the DCI-ALA and ENI budget lines: they have the highest number of fiches (eight out of 13), while more DAs have been signed under the EDF. Findings from the questionnaire The questionnaire to EU Delegations included one question related to this JC: whether clear instructions and guidance have been provided to the EUDs on how to complete the DA and TA assessment fiches (as such, this table applies both to JC 8.3 and JC 8.4). Among 35 respondents, 3% strongly agreed, 29% agreed, 26% disagreed, 3% strongly disagreed and 40% were of no opinion. One reason underscored for the mixed assessment is that there are no comprehensive DA guidelines nor a comprehensive compilation of guidance notes on DA and that at times different guidance notes appear to contradict each other. It was also commented that there is no complete set of templates to be used for the different steps of preparing and implementing a DA (budget table; reports;…).

97


Figure 3.8.4 EUD questionnaire response: Clear instructions and guidance have been provided to the EUDs on how to complete the DA and TA assessment fiches


0% 20% 40% 60% 80%

%

Strongly agree 1 3

Agree 10 29

Disagree 9 26


No opinion 14 40


3.8.7 Quality of TA Assessment Fiches (JC 8.4)

JC 8.4. The TA Assessment Fiches (actually submitted) provide sufficient and relevant information for taking a deliberate and well-motivated decision to accept the transfer of another donor agency

The objective of this JC is to assess the quality of information provided in the TA Assessment Fiches (possibly supported by information from the Action Fiches) about why the transfer from another donor has been accepted. The analysis focuses on the 35 TAs of the 10 case studies. It should be noted that in general the fact that the EU should undertake an assessment to justify why it should receive funding for one of its projects was met with surprise during interviews. Findings from case studies As indicated in the methodological section, the availability of only one TA assessment fiche (related to the PAOSC project in Mali) does not make it possible to assess the information included in those fiches. Action Fiches for TA projects rarely explained the rationale for the EU being a fund managing donor. There is generally little explanation, contrary to most Action Fiches related to DA projects, about the ‘comparative advantage’ of the EU, which would justify the use of TAs. Often, the period between the signature of the Action Fiche and the start of the project was quite long and the Action Fiche did not reflect appropriately the situation. The Action Fiche for the Private Sector Reconstruction project in Palestine was written two years before the TA with Japan was concluded. On the basis of mostly the lack of TA Assessment Fiches, the quality of the decision-making process for TA proposals has been categorised as average. Among the 13 TA projects assessed, the quality of the process was assessed as good in four cases, as average in six cases and as weak in three cases.

98


Figure A.2.8.1 Case study scores: Quality of the decision making process and assessment of TA proposals


Benin 0 2 0 0 2

Ghana 0 1 0 0 1

Mali 0 1 0 0 1


Nicaragua 0 0 0 0 0

Palestine 0 0 0 3 3

Tanzania 0 0 1 0 1


Haiti 0 0 1 0 1


Findings from the questionnaire The table in JC 8.3 presenting the results of the questionnaire on whether clear instructions and guidance have been provided to the EUDs on how to complete the DA and TA assessment fiches is partly relevant for this JC as it doesn’t make the distinction between DA and TA Assessment Fiches.

99


3.9 EQ 9: Quality of cooperation

EQ 9: What has been the scope and quality of the cooperation between the EU, the DC partner and the implementing entity in the partner country during implementation of the project/programme (partly) funded through DC?


This Evaluation Question assesses the scope and quality of cooperation between the different DC actors during the implementation of the DC funded project or programme. The scope and quality of cooperation is assessed in terms of (i) establishment of and adherence to clearly defined rules and procedures, (ii) timely delivery of good quality reports, (iii) frequency and quality of meetings, (iv) role and intensity of monitoring carried out by the EU and (v) timing, quality and follow up of mid-term reviews, final evaluations and audits. Furthermore, the question reviews whether the EU and the DC partner have conducted a review/evaluation of their cooperation in the context of the DC agreement. The different actors are the EU (both HQ and Delegations), the DC partner (both HQ, Development Agencies and offices in partner countries), the implementing entity (in case the EU or DC partner ‘sub-delegates’ or ‘sub-contracts’ the implementation of the work to a third party) and the counterpart entities in the partner country.


The extent to which the DC modality can achieve its objectives depends for a great deal on the quality of the cooperation between all stakeholders during the implementation of the DC-funded project or programme. A clear definition of rules, roles and responsibilities and adherence to these rules are a few of the requisites for a successful implementation of the DC project and as such, contribute to the overall effectiveness and efficiency of DC as a funding modality. This EQ is a process evaluation question related to assessing the quality of the implementation, instead of an impact evaluation question related to measuring the final results of having used the DC instrument. Nevertheless, this EQ is important and relevant because it looks at the process that contributes or hinders the achievement of the outcomes mentioned in the IED, i.e. the EQ assesses whether DC is operationalised in such a way that it can be expected to lead to the desired outcomes. It has to be noted that this EQ does not intend to conduct a full-fledged project evaluation of the respective DC projects or programmes. That is outside the scope of this evaluation. The focus is on coordination, cooperation and reporting issues


JCs. 9.3, 9.4 and 9.5 have been merged as they each deal with operational issues and as the focus of JCs 9.4 and 9.5 is very narrow, and as such does not require to be the subject of a specific JC. For some projects, the evaluation faced information gaps on certain issues, such as final reports and DC project evaluation reports.

100


Findings from relevant studies and evaluations The OECD underlined already in 2003 that “donors involved in delegated co-operation arrangements should have a clear, shared understanding of their respective roles and responsibilities”109. For good cooperation, it is suggested to draft and sign a Memorandum of Understanding. In addition, it is emphasised that the detailed agreements about the delegated cooperation should be made available to the partner country government in order to clarify each other’s roles. The OECD paper stresses furthermore the importance of adhering to the agreed roles and procedures: “The success of delegated co-operation is dependent on the participating donors’ compliance with the agreed roles and procedures reached in a delegation agreement. In particular, a delegating donor must communicate through the lead donor rather than directly to the partner country in the areas of delegation. Failure to do so can impose costs on the partner country and may constrain project implementation and future co-operation efforts”110. The evaluation of the Commission’s cooperation with UN institutions gave examples of which sort of issues as regards the coordination between the EU and other donors, could influence the results of the project or programme. While the UN reporting was quite often delayed, the UN complained that the Commission was a highly demanding donor. Furthermore, both agencies had a different understanding of ‘partnership’ and the level of involvement in day-to-day management in projects111. Similar issues with regard to expectations and misunderstandings could become apparent when evaluating the cooperation between the EU and the other stakeholders involved in DC. Norad conducted a survey in 2010 as regards the Nordic Plus Delegated Cooperation initiative112. This is a joint initiative of the Nordic Plus countries (Denmark, Finland, Ireland, the Netherlands, Norway, Sweden and UK) launched in October 2006 to follow up the commitments made with regard to the international aid effectiveness agenda. The survey used a questionnaire which was sent to the embassies/offices of the participating donors and a couple of HQ departments dealing with the DC Agreements from Denmark, the Netherlands, Norway, UK and Finland. About 70 missions (out of 100) responded to the questionnaire113. 43% of the respondents said that they did not find it difficult to adhere to the agreed roles and responsibilities. However, 40% reported problems encountered in this respect, such as the difficulty to remain a silent donor when either the Lead Donor was not performing or when HQ demanded a more active role. Furthermore, a couple of cases were reported as regards the Lead Donor not complying with the agreed roles and responsibilities vis-à-vis the partner country concerning communication and reporting. Similar practical problems could be discovered when evaluating the DC between the EU and its MS.

109

OECD/DAC, Harmonising donor practices for effective aid delivery, DAC guidelines and reference series, Chapter 6,

2003. 110

Ibidem. 111

Evaluation of Commission’s external cooperation with partner countries through the organisations of the UN Family,

evaluation commissioned by the ED and carried out by ADE, May 2008. 112

NORAD, The Use of the Nordic Plus Practical Guide to Delegated Cooperation, Survey 2010. http. 113

11 were Danish embassies, four were Finnish, 18 were Dutch and 25 were Norwegian embassies. Thirteen of the

respondents were DFID field offices, including a few HQ departments.

101


3.9.4 Cooperation on the basis of clear rules and procedures (JC 9.1)

JC 9.1. The EU, the DC partner, the partner country and the implementing entity cooperate on the basis of clearly defined rules and procedures.

This JC focuses on the rules and procedures of DC at the level of implementation of specific DC funded projects and programmes, because a solid set of rules and procedures improves the chances of a fruitful cooperation. More particularly, this JC deals with assessing the clarity of and the adherence to the rules and procedures. Findings from the desk study In many cases, the DA contracts had to be tailor-made by adding Special Conditions, because the standard DA contract is not adapted to various particular situations and types of DC partners. The standard contract templates of DAs were not in line with the procedures and work methods of some of the DC partners. Therefore, amendments of the General Conditions of the DA agreement were needed and/or additional MoUs had to be prepared and signed. The preparation and negotiation phase of various DAs took therefore quite some time. Another observation from the documentation was that reporting requirements are much more specific for DAs than for TAs. For a DA, the DA partner is required to provide the EU with full information on the implementation of the project. Before signing the DA, the DA partner has to provide a work plan for the first period of implementation. The submission of annual implementation reports and a final report is also obligatory. In a questionnaire circulated in 2012, various EUDs observed that the procedures were often not clear and that the HQ had to be consulted on certain legal aspects of a DA114. This was especially the case in the first years of DC, due to the fact that the instrument was introduced without specific guidelines and training. The Evaluation Team came across anecdotal evidence of lengthy DA negotiations because of ambiguity or disagreement about the conditions. No major problems were identified as regards the preparation of TAs, but this may be this was caused by the fact that some ‘problematic’ TAs were eventually not signed because no agreement could be reached. Two main issues with regard to the procedures are outlined below: 1) DA partners have to comply with EU Financial Regulations According to the DA contracts, a DA partner is allowed to use its own procedures when implementing a DA Action (project). This seems to be straightforward, but the paragraphs above illustrate that the necessity to comply with the EU Financial Regulations can sometimes complicate reaching agreement on which procedures to use. EUDs signalled that also during implementation the use of own procedures by the DA partners may sometimes conflict with the EU procedures, for example in the case of monitoring missions (the EU has its own monitoring missions, which may sometimes interfere with the M&E planning of the DA partner). In that same questionnaire various EUDs indicated also that poor knowledge of and/or misunderstanding about the procedures at the level of the DA partner’s project team had increased their DA-related workload. Another specific example which caused problems during implementation was the D+3 rule. This EU rule requires that all funds have to be committed within three years after signing the Financing Agreement. This rule was not always clear to the DA partner right from the start of implementation and in some cases turned out to be difficult to comply with.

114

Responses to the Questionnaire on DC which was distributed in 2012 to the Directorates, serving as input for the

Report on DC 2007-2012.

102


An example is a DA with BTC in Tanzania, where the Action (project) had to be split in two parts (on paper) to prevent the EU from contributing to a project component that was managed in a way not allowed by the Financial Regulations. A DA in Timor-Leste was confronted with a different interpretation of concepts: the DA partner assumed that the Action was a form of co-financing and that thus the EU rules of nationality and origin did not apply, while the EU considered it as a 100% EU-funded action. The DC questionnaire of 2012 also recalls a case in which an EUD abandoned the idea of a DA after almost one year of preparation due to conflicting implementation rules and procedures115. 2) The issue of sub-delegation During implementation, misunderstandings with regard to procedures and difficulties with adhering to some of the rules were observed. Moreover, procedures were not always clear to both the EUDs and the DC partners. Misunderstandings about regulations surfaced in some cases. It was noticed that DA partners who sub-delegate implementation of their activities to the partner government have most difficulties with agreeing and complying with the procedures, as the standard templates are not aligned with their mode of operation. The Report on DC 2007-2012 points at two issues of confusion among the EU and the DC partners with regard to the General Conditions. The first is the difference between “budget implementation tasks” (i.e. taking decisions on spending EU funds) and “project implementation tasks” (i.e. carrying out project activities such as technical assistance). DC is meant for delegating budget implementation tasks, and not for delegating ‘project implementation’ (for which grant contracts should be signed). The second is a related difference between “sub-delegation” (i.e. delegating budget implementation tasks to an implementing partner) and “sub-contracting” (i.e. delegating project implementation to a service provider). Sub-delegation to a private entity is not allowed according to the EU regulations. However, because of different use of definitions, there have been confusion about whether some arrangements were in fact sub-delegation or sub-contracting. An example is a DA signed with GIZ for a project in the Palestine, where sub-delegation was involved: many stakeholders had to be consulted and there were extensive discussions about the legal terms of sub-delegation. A review of the documentation uploaded in CRIS revealed that there are indeed numerous cases when amendments had to be made to the General Conditions to clarify the points mentioned above. This is often related to the differences in the modus operandi of the various DA partners. For example, KfW and AFD enter into a Financing Agreement with the partner country, which is responsible for the execution of certain projects and programmes. In those cases, a ministry or a public agency becomes responsible for the implementation of the project. Various DAs studied by the Evaluation Team during the desk phase include provisions in the Special Conditions allowing procurement and contracting to be done by the implementing ministry or agency instead of directly by the DA partner. The General Conditions were amended in 2013. One of the new conditions was that in case of sub-delegation the implementing agency had to undergo an additional pillar assessment. In 2014, “sub-delegation” was added as a specific additional pillar to the pillar assessment of DA partners. Other DA partners may hire a private contractor (service provider) for project execution, but remain themselves fully responsible for management of the DA funds. Sometimes some of the DA arrangements were quite close to those of a grant agreement. For these cases, the General Conditions were accepted without many amendments.

115

Responses from Unit E3 covering West and Central Africa, the case refers to the EUD in DRC and DFID.

103


Therefore, as regards clarity and application of rules and procedures, some deficiencies can be observed at two levels: first, between the EUD and the EU HQ and second, between the EU and the DC partner. Most difficulties were faced in the negotiation phase leading to the signing of the DAs. In several interviews, the continuous change of the DC contract templates was mentioned as a costly and time-consuming issue. In addition, in some cases, misunderstandings about regulations surfaced again in the implementation phase. It was noticed that DC partners sub-delegating implementation have most difficulties with agreeing and complying with the procedures, as the standard templates are not aligned with their mode of operation. Findings from case studies The case study analysis has confirmed that some tension has been observed in the implementation of DA projects with respect to the need for DA partners to respect EU Financial Regulations. One particular issue has been the D+3 rule (according to which all funds have to be committed within three years of the signature of the financing agreement) – this was the case in the PAFIRIZ in Benin, the Governance Programme in Timor-Leste and the TVET project in the West Bank. The application of that rule was particularly challenging in the case of DC because the DAs were often signed several months after the signature of the Financing Agreement. It implied that for projects with no co-financing from the DA partners, all contacts had to be signed within a period of often not much more than two years, otherwise funds would be blocked. Another reported source of difficulty with regards to procedures, especially in the early days of DC, has concerned the application of EU eligibility rules in DA projects, especially as the definition of the ‘action’ was not standardised and often included non-EU funds In interviews, the frequent changes in DC contract templates and regulations were also mentioned as a time-consuming issue, which often led to misunderstandings. Various EUDs have also indicated that limited knowledge of and/or poor misunderstanding about EU rules and procedures have been frequent and have increased their workload. Finally, in the process of the formulation of DAs, some tension was sometimes observed between the Commission HQ and EUDs. All stakeholders, including EUDs and the TA and DA partners, have indicated that there is still a lack of clarity regarding the application of rules and procedures. The field missions also confirmed the challenges with regard to the DA partners which sub-delegate implementation to a third party. The case studies have illustrated a few examples of the (legal) difficulties which occurred when a DA was signed with a DA partner which itself does not implement the project, but instead sub-delegates implementation (to a government agency). One issue that has arisen in this case is the question of the ‘liability’. This was for example a sensitive point in the case of the formulation of the DA in the water sector in Benin concluded by the EU and the Embassy of the Netherlands. This was solved by a MoU annexed to the DA and signed by the Government of Benin, the EUD, and the Embassy of Netherlands, stipulating that the fiduciary responsibility would lie in the hands of the Government of Benin. With regards to TAs, the view from TA partners was often that too much discretion was left to the Commission in the management of TA projects, reflecting the fact that TAs were very lean documents, compared to DAs. The General Conditions in TAs are less specific than those of DAs. As exemplified by the PAOSC TA in Mali, the Commission sometimes did not demonstrate flexibility in taking on board specific requirements of TA partners. Sometimes, the EUD itself lacked autonomy to address issues and make decisions, having to refer to Commission HQ, which often led to delays in decision-making or very rigid/formalistic decision-making (such as when requesting interest payment to TA partners following delays in tranche payments). This then created tensions on the ground. Some DC partners have viewed this dichotomy between the length and requirements in

104


respectively DAs and TAs as reflecting a lack of ‘reciprocity’ between the EU and DC partners in the management of DC. Finally, the case studies have underscored that partner countries have little knowledge about the details of DAs and TAs. When signing the Financing Agreement of a project, the partner country agrees with using the DC modality for funding the project or programme (because this is mentioned in the Financing Agreement), but it appears from the case studies that partner countries had little awareness of the details of the DA or TA, which are in essence contracts between DC partners. Findings from the questionnaire In the questionnaire, the EUDs have also been asked to assess whether rules and procedures as regards the cooperation between the EU, the DC partner and - if applicable - the implementing entity in the partner country were clearly defined. Among 35 respondents, 9% strongly agreed, 71% agreed, and 20% disagreed. These findings show a result of satisfaction about the clarity of DC procedures that is more positive to what the case studies have underscored. This partly reflects the fact that, in the case studies analysed, it is mostly other DC partners, less the EU, who underlined the problems described. Figure 3.9.1 EUD questionnaire response: Rules and procedures as regards the cooperation between the EU, the DC partner and - if applicable - the implementing entity in the partner country were clearly defined


0% 20% 40% 60% 80%

%


2 Agree 25 71

3 Disagree 7 20


5 No opinion 0 0


3.9.5 Provision of timely and adequate information (JC 9.2)

JC 9.2. The fund managing donor has provided timely and adequate technical and financial information on the implementation of the project/programme.

Introduction This JC focuses on assessing whether the fund managing donor, which is the EU for TAs and a DA partner for DAs, has provided timely good quality implementation reports to the fund transferring partner (= the delegating donor)116. Timely is interpreted as in line with the agreed time schedule; good quality is measured in terms of complying with the reporting guidelines as stipulated in the General Conditions (or Special Conditions in case of TAs) and the expectations of the DC partner.

116

In case the project/programme is implemented by a third party (the implementing entity), that entity will most likely

produce the annual implementation reports and work plans, but the fund managing donor still has the responsibility to

assure that those documents are produced and transmitted to the delegating donor.

105


In the General Conditions of a DA, it is stipulated that annual implementation reports should be submitted to the EU within 60 days after the final date of the reporting period. It is also indicated that the scope of the report has to cover the whole Action (project) regardless whether the Action is fully financed or co-financed by the Commission. The Conditions provide even a template for those annual reports (see box below). Template for the annual Implementation Reports of a DA-funded action (project)

A similar outline for the Final Report is also included in the General Conditions (see box below). That report has to be submitted within 6 months after the project/programme has ended. Template for the Final Report of a DA-funded action (project)

The guidelines on reporting have been changed over the years. The extracts (see boxes above) are taken from the General Conditions as introduced in 2013117. Previous versions of the General Conditions also included instructions on the Implementation Report, while

117

Copied from the DEVCO Companion version 4.1, September 2014.

The Implementation Report shall at least include: (a) Summary and context of the Action; (b) Summary of how each activity was implemented and information on the measures taken to

identify the European Union as source of financing, including information on expenditures actually committed and incurred per activity;

(c) Difficulties encountered and measures taken to overcome problems; (d) Changes introduced in implementation; (e) Achievements/results by using the indicators included in this Agreement; (f) Information on the control and audit measures to which the Delegatee body itself has been

subjected, as well as measures of control carried out on Implementing partner(s). In case where an inspection was carried out a summary inspection report indicating the number and results of on-the-spot checks audits already carried out;

(g) Overall financial statement providing an account of the use of the funds transferred to the Delegatee body;

(h) Work plan for the following period including objectives and indicators of achievement. If the report is sent after the end of the period covered by the preceding work plan, a new work plan, albeit provisional, is always required before such date;

(i) When appropriate, a request for payment.

Source: General Conditions of the Delegation Agreement, Article 2.6

The Final Report shall at least include: (a) a full summary of the Action's income and contributions received and costs incurred,

including the total amount actually paid by the Delegatee body or Implementing partner(s) to third parties, the total amount of the financial compensation for the administrative costs and the total of the real costs of the technical assistance provided by the Delegatee body to third parties thereby enabling to determine the total balance to be recovered by the Commission from the Delegatee body;

(b) the control and audit measures to which the Delegatee body itself has been or will be subjected including Implementing partner(s), accompanied, if appropriate, by current information on measures taken to remedy any identified problems;

(c) a summary table setting out the final amount of each grant and/or procurement contract; (d) description of the results generated by each operation and their impact in relation to the

objectives of the Action using the indicators included in the Agreement; (e) details of transfers of assets mentioned in Article 7.3 if relevant; (f) when appropriate, a request for payment. Source: General Conditions to the Delegation Agreement, Article 2.8.

106


much more emphasis was put on financial reporting, e.g. a financial statement on the use of the administration fee was demanded. The current reporting guidelines request more information about the technical implementation of the project, while also a work plan has to be included. The guidelines for the Final Report remained more or less the same over the years. Interestingly, the General Conditions of the TAs do not contain reporting guidelines. There is only a reference to the Special Conditions, where arrangements can be made as regards type and frequency of the reports. Findings from case studies The case studies have underscored that technical and financial reports, including final reports, have been consistently produced by DA partners, and that those were generally timely. The quality and scope of the reports however appears to have varied significantly. In a number of cases (such as Ghana), despite receiving regular progress and annual reports, including financial statements, EUD staff complained that it was difficult to obtain a good picture of the funded activities. The EU was also not always satisfied with the level of detail of the reports. Many problems were observed with the agreements where reporting is done by a ‘third party’: in the case of KfW for example, the Project Executing Agency is responsible for reporting, which often outsources this to a hired consulting firm. There are also examples of DA projects where AFD sub-delegated reporting to a third party. Also, contrary to the requirements of article 2.6, reports were not systematically covering the whole action (sometimes reflecting a lack of clarity on what the action was). One example is the Programme d’appui à la déconcentration et au développement communal (PDDC) in Benin, for which a specific report was produced for the component of the PDDC financed by the EU. The reporting requirements as regards the TAs were in most cases specified in the Special Conditions, which led to more flexibility and discretion in the way the EU would report to TA partners. For the relatively small TAs contributing to the PEGASE programme in the Palestine, reporting requirements were limited to a copy of the payment report. For larger TAs, more extensive reports were often prepared (e.g. for a TA concluded with DFID in Tanzania, quarterly visits and reports were demanded). In Haiti on the other hand, the AFD complained that the reports produced by the EUD for the APONTRA project, for which its TA was contributing € 39 million, were particularly thin. Reporting was relatively problematic in the case of the GCCA. The GCCA management at EU HQ had to inform the TA partners about project implementation, but it was dependent on information to be obtained from the EU Delegation, which in turn was waiting for information from the DA partner. This chain of information with various ‘transmission stations’ impacted on the timeliness and sometimes also on the quality of information finally provided to the TA partner. Findings from questionnaire In the questionnaire, the EUDs have been asked to assess whether in the case of DAs, the DC partner has provided timely adequate technical and financial information on the implementation of DA-funded project(s)/programme. Among 35 respondents, 3% strongly agreed, 60% agreed, 29% disagreed and 9% had no opinion. Such results appear broadly in line with the findings from the case studies.

107


Figure 3.9.2 EUD questionnaire response: In case of DAs, the DC partner has provided timely adequate technical and financial information on the implementation of DA-funded project(s)/programme(s)


0% 20% 40% 60% 80%

%


2 Agree 21 60

3 Disagree 10 29


5 No opinion 3 9


In the questionnaire, the EUDs have also been asked to assess whether agreed rules and procedures were adhered to by the EU, the DC partner and - if applicable - the implementing entity in the partner countries (e.g. reports were delivered on time, the quality of the reports was satisfactory, etc.). Among 35 respondents, 9% strongly agreed, 71% agreed, 17% disagreed and 3% strongly disagreed. Those results are very similar to the answers to the previous question. Some respondents expressed the view that DC partners and in particular the fund managing donor should not just transfer the reports produced by the implementing entity but also ensure proper follow-up and monitoring of programme's implementation and quality review of documents. Figure 3.9.3 EUD questionnaire response: Agreed rules and procedures were adhered to by the EU, the DC partner and - if applicable - the implementing entity in the partner countries (e.g. reports were delivered on time, the quality of the reports was satisfactory, etc.)


0% 20% 40% 60% 80%

%


2 Agree 25 71

3 Disagree 6 17


5 No opinion 0 0


3.9.6 Quality of the coordination between the partners (JC 9.3, 9.4 & 9.5)

JC 9.3. The coordination between the DC partners has been satisfactory.

JC 9.4. The DC partner has organised a mid-term review and/or final evaluation of the project/programme concerned.

JC 9.5. The DC partners have reviewed and/or evaluated their cooperation in the context of the DC agreement.

108


These JCs, which have been merged because they focus on related operational issues, assess mostly the quality of coordination between DC partners. They include an analysis of the frequency and content of both formal and informal coordination meetings and of the coordination of the DC project’s evaluations/monitoring processes. Most inputs for answering these JCs have been collected during the case study phase. Findings from the desk study In one of the responses to the Questionnaire on DC circulated in 2012118, it was noted that EU monitoring missions often discovered insufficient coordination between the major stakeholders of DA-funded projects and insisted on improving the dialogue between the donors and the project implementation team. It was also indicated that the relation between the EUD and the DA partner can sometimes be a little uneasy, because the EUDs have to validate whether the DA partner respects the Financial Regulations and fulfils the visibility requirements. As explained in JC 9.1, specific EU rules are sometimes not clear to the DA partner, which can cause frictions during implementation, e.g. when the DA partner has to comply with certain disbursement restrictions. When going through the available documentation of the DAs and TAs of the case studies, anecdotal evidence was found as regards coordination shortcomings on case of a project supported by two DAs, which caused a series of request for clarification from the partner country. Another DA file included a separate MoU between the EU and the DA partner about informal working arrangements in the country where the project was implemented. As this was a DA supporting a second phase of a project, this could refer to difficulties being experienced during the implementation of the DA having supported the first phase of the project. Another response to the DC Questionnaire of 2012 cites a case where the EUD communicated directly with the sub-contractor instead of via the DA partner. Some of the documentation on DAs as well as results from a questionnaire circulated at the level of EU Directorates and EUDs in 2012 suggested unease in the coordination between EUDs and DC partners in some cases. Specific EC rules were sometimes not clear to the DA partner, which can cause frictions during implementation, e.g. when the DA partner has to comply with certain disbursement restrictions. Findings from case studies The quality of the cooperation between the EU and the DC partner during implementation of the DC agreements was on average good. In the following table, the scores regarding 45 projects and programmes are summarized regarding the quality of the cooperation. The table shows that in 12 projects the quality of that cooperation was assessed as very good, in 21 projects it was assessed as good, in 11 it was average, and in one it was considered weak. That latter project is the TECNICA project in Nicaragua in which AECID lamented that it was treated more like a supplier of technical assistance than a real partner in cooperation. The quality of the cooperation was generally seen as slightly higher in the case of DAs than TAs. The same table underscores that, while on average the quality of cooperation in TA projects was good (50% scored ‘good’), the percentage of TA projects scoring ‘very good’ (16%) was quite lower than for DA projects (30%).

118

Food facility.

109


Table 3.9.1 Case study scores: Quality of the cooperation between the EU and the DC partner during implementation of the DC agreement

Note: a TA in Tanzania was not scored because implementation has not started.

In the case of DAs, the broadly positive scores reflect the fact that while implementation challenges occurred (such as those outlined in JC 9.1), there was in general a good level of coordination and cooperation among DA partners. Coordination meetings took place regularly, both formal and informal. There are no examples of projects were there was a sustained lack of coordination or cooperation. When implementation problems occurred, dialogue between the EUD and the DA partner took place in most cases, and problems were generally solved. Given the relative lack of guidance on how DAs should be managed on the ground, the specificities of the coordination and cooperation varied however, and largely depended on the project managers at the EUD and the DC partners’ agencies on the ground. As such, the cooperation approach taken could vary significantly from one project to another and even within one project (in case of a change of staffing, such as in the case of the PAFIRIZ in Benin). In some DA projects, the EUD took a relatively ‘silent partner’ approach while in others, it was much more active, in particular in the policy dialogue. As the case studies have underlined, there was in practice not one model of cooperation/delegation. With regards to TA projects, coordination was broadly satisfactory, but the different nature of DC compared to DAs and reciprocity issues may have contributed to the slightly less positive scores. TA partners have also co-financed a smaller amount and a lower percentage to the total funding of a TA project than the EU has done when funding a DA project. The average size of a DA has been €7.7 million and for a TA €4.9 million (see table 2.2 in chapter 2). Therefore, the EUDs may have been less inclined to actively cooperate than Member States have in the case of DAs. A number of TA partners has expressed the view that much discretion has been left to the EU in the management of TA projects, reflecting the fact that TAs have been very lean documents, compared to DAs. The General Conditions of TAs have also been less specific than those of DAs. In some cases, the EUD itself has lacked autonomy to address issues and make decisions, having to refer to the EU HQ. This has often led to delays in decision-making or quite rigid/formalistic decision-making. Concerning the coordination of monitoring and evaluations of the DA projects, it was observed that a few EU monitoring missions were carried out, which were not always fully coordinated with the DA partner, but that this was not really a subject of tensions between


Benin 1 5 1 0 7

Ghana 0 0 3 0 3

Mali 1 2 1 0 4


Nicaragua 1 2 0 1 4


Tanzania 0 2 0 0 2


Haiti 0 0 3 0 3




110


DC partners. According to Article 8 of the General Conditions of DAs, “representatives of the Commission shall be invited to participate in the main monitoring and in the evaluation missions relating to the performance of the Action. The results of such missions shall be reported to the Commission by the Delegatee body.” In addition, the EU can carry out its own monitoring and evaluation missions. In the case of TAs, a number of mid-term and final evaluations were carried out by the EU. The review of some of those evaluations suggests some involvement of outside stakeholders, including TA partners, in those evaluations. For TAs, the description in the General Conditions as regards the participation of the TA partner in monitoring and evaluation missions is a bit less strong than for DAs. Article 11 of the General Conditions of TAs states that “representatives of the Donor may be invited to participate in the main monitoring and in the evaluation missions relating to the performance of the Action.” The word “shall” used in the General Conditions of DAs is replaced by “may” in the General Conditions of TAs. Finally, no strategic evaluations or reviews of DC by DC partners have been identified within the case study countries or at a broader level. Some evaluations of DC projects referred briefly to the associated DC modality, but not in strategic terms. As the DC modality has now been implemented during a number of years and as the instrument is seen as an interesting tool in the aid modality mix among many DC partners, several DC partners appeared to share a real interest in the evaluation of the DC instrument carried out by the evaluation team. Findings from questionnaire In the questionnaire, the EUDs have been asked to assess whether coordination between the EUD, the DC partner and - if applicable - the implementing entity in the partner country has been satisfactory. Among 35 respondents, 11% strongly agreed, 77% agreed, and 11% disagreed. Those results are broadly consistent with the findings from case studies. Figure 3.9.4 EUD questionnaire response: The coordination between the EUD, the DC partner and - if applicable - the implementing entity in the partner country has been satisfactory.


0% 20% 40% 60% 80%

%


2 Agree 27 77

3 Disagree 4 11


5 No opinion 0 0


111


evaluation of delegated cooperation desk report · evaluation of the eu aid delivery mechanism of...

Documents